Every marketer I talk to right now has more tools than they’ve ever had in their career. Websites you can stand up in an afternoon. Free reach to anyone on earth through social. Now AI sitting on top of all of it, writing the copy, building the images, answering in seconds what used to take a week.

And the needle is not moving.

Not because the tools are bad. Not because the team isn’t trying hard enough. It’s because tools amplify whatever’s underneath them, and most of the time what’s underneath isn’t solid enough to amplify.

I’ve been circling this problem longer than I realized. In 1998 I was working nights at a bank, running mainframe jobs on a Unix box, balancing transactions while the system chewed through its batch. There was a lot of downtime in that job, and I filled it writing. Not marketing. Just sitting with a hard question long enough to actually get somewhere with it, instead of reaching for the first answer. I never finished college. I didn’t know it at the time, but that was the first real practice I had at digging past the obvious answer to the true one.

Tom Nixon, my co-host on Bullhorns & Bullseyes, has been circling the same instinct from a different direction for thirty years, across two novels and a career in advertising production. Neither of us invented this argument last week. We just finally said it out loud on air: we’re writing a book together. It’s called The Marketing Reformation.

None of that is the point. The point is the argument underneath it. Every time an industry’s quality gate comes down, volume explodes and the average collapses. Record labels used to force a musician through years of proving they could actually play before anyone heard them. That gate came down, and there’s a line from Toto’s Steve Lukather that Tom’s heard attributed to him more than once: back then you needed real musicianship to make a world-class album, now all you need is a celebrity and a laptop. Advertising had the same gate. Mad Men-era agencies were full of their own sins, but at minimum they made someone do their homework before an idea reached a client. WordPress took the same kind of gate down for building a website. You don’t need a developer anymore, so you don’t get one, and now there are a hundred million websites that say nothing. AI is doing this to marketing right now, faster than any of the previous collapses.

Tom put words to the distinction we keep circling back to on air: anyone can come up with an idea. The job was never coming up with it. The job is making the idea hold with the customer. Marketing’s default move is skipping straight to the coming-up-with part and calling it strategy.

I wrote down a line from a Brian Clark newsletter a while back that’s stuck with me since: once you understand your wiring, you don’t choose your movement, you come to recognize it, because you understand which movements already resonate with who you actually are. That’s what this book is for me. Not a new idea. A recognition of one I’ve been circling since a bank mainframe in 1998.

Tom’s friend Chuck Bean has a line he uses with clients constantly: you’re not arguing with me, you’re arguing with the data. That’s the whole discipline in one sentence. Find the customer’s actual truth first. Then, and only then, turn the amplifier on.

I was raised Lutheran, and the title isn’t an accident. The Reformation wasn’t Luther inventing a new belief. It was a correction back to what people already, authentically believed, once the gatekeeper standing between them and it got out of the way. That’s what Tom and I think marketing needs right now: a correction back to the truth the whole apparatus was supposed to serve, before anyone turns the amplifier on.

The prologue is written. The outline is locked. We’ve already got three or four chapters down, passed back and forth between us. We said out loud on air that we’re aiming for this year. Now we can’t renege.

We’ll be publishing the deeper cuts here as we write. If you want to watch this get built instead of just reading the finished book, this is where that happens.

Full episode is up now. Next week, Aimee Schuster joins us. Catch Bullhorns & Bullseyes on YouTube, Apple Podcasts, Spotify, and wherever else you listen.

WATCH NOW »

I was sitting in a brand-new podcast production studio recently. Beautiful room. Two cameras. The kind of operation built for CEOs who want to show up on YouTube and look the part.

The owner told me about two of his clients. Same studio. Same gear. Same production cadence. Completely different relationships with the work.

The first client has four episodes published. Forty views on each one. He hasn’t checked his subscriber count in months. He just closed two new deals directly from those four episodes. The deals paid for the entire year of production. He’s already in the black. He’s not refreshing his analytics dashboard. He doesn’t care.

The second client has 1,200 subscribers and is upset. Not about pipeline. About the subscriber count. He wants the number to go up. He’s three quarters in and hasn’t closed a deal from the show.

Forty views. Two deals. Production paid for.

Twelve hundred subscribers. Zero deals. Frustrated.

Which one of those CEOs is doing marketing?

The follower count is not the asset.

Most companies think they’re building an audience. They’re not. They’re renting one.

You have 15,000 LinkedIn followers. LinkedIn changed its algorithm on a Tuesday. By Wednesday, your reach dropped 60%. You didn’t do anything wrong. The platform decided to show your content to fewer people. That’s it.

A follower count is permission to appear in a feed. The platform decides when, how often, and to whom. That’s not an audience. That’s an audience on loan.

For Season 3, Episode 6 of Bullhorns & Bullseyes, Tom and I brought back Brian Clark to make the operational distinction sharper than I’ve ever heard it made.

If you don’t know Brian, he practically coined the practice of content marketing before the term existed. Built Copyblogger from a one-man blog into a business with 65 people doing eight figures. Sold it. Walked away. Started Further as a personal newsletter that became the most thoughtful thing in my inbox every Sunday for years before I ever met him.

He gave us the test. One question. Two parts.

Can you take the list with you? Can you take the content?

If both yes, you own it. If one no, you’re a tenant.

By that test, your 15,000 LinkedIn followers are LinkedIn’s audience. Your 50,000 Instagram followers are Meta’s audience. Your subscribers on YouTube are Google’s audience. The platforms let you talk to those people on terms the platforms set. Until they don’t.

The list you can export is the asset. Everything else is borrowed attention with an expiration date.

The CFO Argument

Here is what I keep running into. Operators read what I just wrote and nod. Then they go to their CFO and ask for budget to build an owned audience. The CFO looks at the cost-per-acquisition column in the paid-media dashboard. Says no.

So I asked Brian, directly, on tape: How would you make this argument to a CFO?

His first answer was a joke. “Tell them to go count beans.”

Then he got serious.

“I would lead with conversion rates of email versus social media. Forty times higher conversions. Social media is not for pitching. Email after sufficient warm-up is. It converts at a much higher rate.”

That 40x figure is not Brian’s intuition. It traces to seminal McKinsey & Company research showing email acquires up to 40 times more customers than Facebook and Twitter combined. Brian was citing established data when he said it. We just don’t talk about it in those terms in marketing conversations anymore, because everyone has gotten used to talking about social like it is the primary channel. It isn’t. It never was.

Tom shared a real client number from one of our retainers. Six thousand six hundred people on the list. A 52% open rate. That means that some significant portion of 3,300 people read the latest send. The same client’s paid ad campaign last quarter had a 12% audience penetration. Excellent by industry standards. Still, it’s no 52%.

The math an actual CFO will respect: Paid acquisition is a variable cost. You stop paying, you stop reaching. Every customer you acquire today resets the clock to zero tomorrow. The cost per acquisition compounds upward every year as more competitors enter the auction.

Owned audience is an appreciating asset. Marginal communication cost approaches zero after the list is built. Every subscriber added is permanent reach. The relationship deepens with every send. The asset is on the balance sheet whether your bookkeeper writes it down or not.

Paid acquisition resets to zero every morning. Owned audience compounds.

That is not a brand argument. That is a balance-sheet argument.

Why the “40- Views CEO” is Winning

Curtis made the observation that put the studio story in focus.

“Forty people watch a podcast and two people buy from you. That’s a pretty good conversion rate.”

It is. Five percent. A direct-mail copywriter from 1965 would call that an excellent campaign.

But the 5% conversion rate is the surface fact. The deeper fact is that those 40 people showed up specifically for that CEO. Not for the category. For him. They recognized something in his judgment they could not get from someone else. Two of them had a problem he could solve. They bought.

The 1,200-subscriber CEO has more reach. He has zero conversions, because his content is the average of his category. It’s AI-assisted, professionally produced…and indistinguishable from any of the other ten shows competing in the same niche. The reach is real. The recognition is not.

Volume is not the problem. The trust is the problem.

The Movement Movement

Around minute forty of the conversation, Brian said something that I think is the load-bearing claim of this entire episode. He was talking about his career. Legal services (because that’s what was available). A virtual real estate brokerage (because it paid well). Then Copyblogger.

The Copyblogger pivot came when he asked himself, for the first time, what kind of business do I actually want? Not what makes the most money. Not what is most prestigious. Not what is easiest to explain to his mother. What kind of work fits who he actually is — his temperament, his soft skills, his way of looking at the world.

He picked a model. He executed against it. Eight figures.

Then he said the thing…

“I always end up starting businesses that are aspects of movements. And movements bond people together strongly because all of you think the status quo is wrong and you want to change it.”

Movements outlast trends. Trends die when the next shiny thing arrives. Movements survive because the people inside them believe the current order is wrong. The audience that compounds is a movement. The follower count is a trend.

If you’re trying to build an owned audience and the work isn’t compounding, the most likely problem is not your email tool. It’s not your cadence. It’s not your subject lines. It’s that you have not yet named what you believe the status quo gets wrong. Until you do, your content is just commentary on the category. Once you do, the people who share that belief recognize you and stay.

Belonging belongs in the hierarchy of needs.

Brian closed the episode with the most quotable single thing anyone has said on this podcast all season.

He was answering a different question. We had been talking about Steve Jobs refusing focus groups. About being yourself and attracting people like you. About the line he keeps writing about midlife and the longevity economy, because that is who he is and that is the audience that has assembled around him.

Then he zoomed out.

“Belonging is such an important part of the hierarchy of needs that it should probably be equated with food and shelter. Without it, people die earlier, they commit suicide. It is so powerful. And now we’re entering this age of AI and everything’s artificial and we don’t know what’s real. What’s going to be coveted even more? Belonging with like-minded people. We’re tribal. That used to be based on appearance or geography. Now it’s identity. What is identity? It’s a bundle of beliefs.”

Identity is a bundle of beliefs. You cannot make it without standing for something, without saying what you believe, and without inviting others to belong with you.

The owned audience is not just a distribution mechanism. It is the infrastructure that lets a community of belief assemble around you. The portability test is the technical definition. The belief assembly is what makes the asset appreciate.

This is what Mark Schaefer was getting at in Episode 5 when he said belonging is the moat AI cannot replicate. Brian just told us what the moat is built on. A bundle of beliefs. An identity. A movement.

AI is not the solution. AI is the amplifier.

Toward the end, Brian said the line that I’ve been working on for the whole season, in his own words and arrived at from a different door.

“The AI stuff is fantastic. But it’s not the solution. It’s an auxiliary concept. It’s your collaborator. It’s your augmentation. And I’m afraid that people are getting this backwards. They’re trying to lead with AI, and they don’t have anything else going on that they understand about themselves.”

I converted it on tape.

“AI is the false prophet. But it can be the amplifier. If you do the work first — which is you — and you codify that, now AI can be the amplifier. The mistake everybody’s making is going to the AI first. Like a prompt is going to save them. Like a blueprint is going to save them.”

AI does not produce belonging. AI does not produce identity. AI does not produce the conviction that builds a movement. AI produces the statistical average of every existing piece of category content. Run that through a rented platform algorithm and the signal degrades twice.

But put AI in front of an operator who has done the work — who knows what they believe, who knows who their audience is, who has built an owned list of people who recognize them specifically — and AI compounds the asset. It writes faster. It tests more variations. It surfaces patterns in the response data. It scales the relationship without thinning it.

The work has to come first. Then AI multiplies it. If you skip the work and lead with AI, AI accelerates your disappearance into the category average. If you do the work and let AI amplify it, AI accelerates your distinction from everyone else in the category.

Same tool. Opposite outcomes. The variable is what you built before you turned the tool on.

The Sovereign Startup

Brian closed by naming his current project. It is called Sovereign Startups. It is for Gen X founders who are watching their corporate jobs disappear into AI-driven layoffs and reorgs, who have the experience and the savings to bootstrap, and who are designing businesses that fit who they actually are instead of chasing the latest formula.

“The perfect blueprint for you is one you design based on your temperament, your skills, your particular expertise, your purpose, what you believe in. The human layer is more important. The AI stuff is fantastic, but it’s not the solution.”

I’ve spent the last year building a system that helps marketing leaders excavate that human layer and codify it for AI. We call it the Brand Guardian. It is not coincidence that Brian and I arrived at the same conclusion through different doors. He came in through audience. I came in through data. The conclusion is the same.

The audience you own is the only audience that compounds. The trust they have in you is the only thing AI cannot manufacture. Build the infrastructure that lets that trust appreciate, and design the business around who you actually are.

That is the only marketing left worth doing.

The Curtis Test

Before you spend another dollar on paid acquisition, run these three on whatever you’re building:

1. Can you take the list with you? If you stopped paying every platform tomorrow, how many humans could you reach directly? That number is your real audience.

2. Of those people, how many would you say chose you specifically? Not because an algorithm served you to them. Because they recognized something in your judgment they cannot get elsewhere.

3. What status quo do you believe is wrong? If you cannot answer that question in a sentence, you do not yet have a movement to build an audience around. You have content. Content does not compound.

If those three numbers are smaller than you’d like, that’s the diagnostic. That is what to fix first.

Next episode, we close the loop on measurement with Aimee Schuster. We’ve now built five of the six layers. Psychology. Data. Story. Community. Audience. The sixth—measurement—is the one your CFO will sign off on.

Tom said it best on tape: “Lesson seven is the one you can invite your CFO to.”

See you then.

— Curtis

P.S. Brian’s two projects worth your time: Further and Sovereign Startups.

P.P.S. If you want to go deeper on the AI-amplifier-not-prophet argument, the March 2 manifesto is here: AI Isn’t Your Prophet. It Might Be Your Amplifier.

P.P.P.S. Listen to or watch the episode wherever you get your podcasts, on YouTube, or right here on Substack.

Drawing on two decades of building businesses on owned infrastructure, Brian introduces the portability test (can you take the list with you? can you take the content?) and explains why a story of forty views and two closed deals beats 1,200 subscribers and zero conversions every time.

The conversation covers the conversion economics of email versus social, why movements outlast trends, how belonging sits beside food and shelter in the human hierarchy of needs, and why AI is the amplifier—not the solution—for operators who have first done the work of understanding who they actually are.

N.B.:

• Read Brian’s newsletter, Further, at further.net.

• Learn more about Sovereign Startups at sovereignstartups.com.

• Connect with Brian on LinkedIn.

Takeaways:

• The portability test: if you can’t take the list with you and the content with you, you don’t own the audience—you’re a tenant.

• A follower count is permission to appear in a feed; the platform decides when, how often, and to whom. An owned list is an asset on the balance sheet.

• Forty views and two closed deals beats 1,200 subscribers and zero conversions. Volume is not the problem—trust is.

• Email converts at roughly 40 times the rate of social media. The channel most teams treat as secondary is the one that actually closes.

• Paid acquisition resets to zero every morning. Owned audience compounds. Those are not equivalent strategies.

• Movements outlast trends because the people inside them believe the current order is wrong—that conviction is what makes an audience stay and grow.

• Belonging sits beside food and shelter in the human hierarchy of needs. In the age of AI, it will only become more coveted.

• Identity is a bundle of beliefs. The owned audience is the infrastructure that lets a community of belief assemble around you.

• AI is the amplifier, not the solution. Fed an operator who has done the work, it compounds distinction. Fed one who hasn’t, it accelerates disappearance into the category average.

• The Sovereign Startup: the best business blueprint is one designed around your temperament, your skills, and what you actually believe—not the latest formula.

Find and Follow:

• Find all episodes at bullhornsbullseyes.com.

• Be sure to subscribe to our Substack to never miss an episode!

• Follow the show on LinkedIn!

• Learn more about Collideascope and Creative Mill at their respective websites.

• Connect with Curtis and Tom on LinkedIn.

• Check out our newsletter, Amplify and Aim!

“Where do you want to get to?”

Lewis Carroll posed this question in 1865, through a Cheshire Cat perched in a tree.

Alice, lost in Wonderland, asks which way she ought to go. The Cat’s response is immediate: “That depends a good deal on where you want to get to.” Alice says she doesn’t care much where, as long as she gets somewhere. “Oh, you’re sure to do that,” the Cat says, “if you only walk long enough.”

Or fast enough, I might add.

If you were lost on a road trip, what would you do? Hit the accelerator to get “wherever” faster? Or would you slow down? Perhaps stop entirely. Ask a local for directions. Consult the GPS. Figure out first where you’re going, know the best way to get there, then proceed with confidence, right?

Unfortunately, what I see most companies and marketing teams doing today, now that they have access to AI, is precisely the opposite. They are chasing Alice’s folly.

“We don’t know where we’re going or how to get there, but let’s hit the gas and get there faster!”

The Map is the Source of Truth, Not Your “Impeccable Sense of Direction”

“Where do we want to get to?”

Most marketing plans never ask—or answer—this most important question.

Not “What’s our budget?”
Not “Which channels are we using?”
Not “What does the content calendar look like?”

Those are real questions, and they get answered early and often. But the question that precedes all of them—the one the whole plan depends on—frequently gets skipped, dressed up in vague performative language, or deferred entirely until the next planning cycle.

It’s easy to read the Cat as clever and Alice as naïve. The more uncomfortable reading is that the Cat isn’t being sarcastic. He’s being precise. If you have no destination, then any road is a valid pathway. The logic is airtight. The problem is that it describes the marketing strategy of most businesses operating in the modern marketing ecosystem.

Measuring Miles Per Hour, Not Progress

Here is what happens when a marketing plan begins without a destination: it fills the gap with measurement.

Activity gets tracked. Clicks, impressions, open rates, follower counts, cost-per-lead. The reporting gets more sophisticated over time. Dashboards proliferate. Somebody books a quarterly review to go through the numbers.

And every quarter, the answer to, “Is this working?” is the same: Compared to what?

Marketing measurement is only meaningful when it’s calibrated against a destination. Without one, you’re measuring the speed of the car, not whether you’re getting closer to your desired endpoint. You’re covering ground—impressions are up, traffic is healthy, the content calendar is running on schedule—but you have no way to know if any of it is moving you anywhere useful. You only know you’re moving.

This isn’t a technology problem. It’s a direction problem. And it cannot be solved with better reporting.

Strategy Is Not a Synonym for Planning

The word “strategy” does a lot of impersonation work in marketing. Plans get called strategies. Tactics get called strategies. Calendars get called a strategy. A documented list of channels with associated budgets gets called a strategy.

A strategy is a theory of how you win. It requires a definition of winning. Without that definition, you don’t have a strategy; you have a plan with no objective attached to it. Rigorously executed, admirably formatted, and functionally directionless.

The Cheshire Cat’s logic is pitiless on this point. If you haven’t defined where you want to get to, you cannot evaluate whether the road you’re on is a good one, let alone the right one. Every path looks equally valid. Every tactic is a reasonable idea. Every new channel is worth exploring. The plan expands to fill the available budget, the team stays busy, and the question of whether any of it is working remains permanently deferred. The car is speeding ahead, after all.

This is not a failure of execution. It is a failure that precedes execution by several months.

The Destination Has to Be Defined and Definitive

Here’s where the parable gets instructive, rather than just diagnostic.

Alice’s mistake wasn’t that she didn’t know where she was going. She was lost in Wonderland. That’s a reasonable state. Her mistake was accepting that not caring about the destination was fine, as long as she eventually got somewhere.

Most marketing “plans” make the same concession. The objective gets written in a form that sounds directional but isn’t:

“Increase brand awareness.”
“Build our online presence.”
“Grow our audience.”
“Generate leads.”

These are descriptions of motion, not destinations. They’re Alice saying, “Any road will do.”

A real destination is specific enough that you’ll know when you’ve arrived…and specific enough that you’ll know if you’re heading the wrong way.

Not “increase brand awareness” but, “Within 18 months, increase our marketing efficiency ratio from X to Y, resulting in an X% increase in sales (or profit margin).” That’s a destination. You can look at what you’re doing and ask whether it points in that direction.

The Cheshire Cat would have nothing to add in such a scenario. There’s nothing left to ask. The cat’s own tongue would be “got,” for once.

Walk Long Enough, in Any Direction, and You’ll Get There

There’s one more line in the Carroll exchange worth sitting with. After Alice argues that she wants to get somewhere, the Cat says she’s sure to manage that, if she only walks long enough.

It’s the most devastating part of the passage, and the most misread. People take it as a kind of optimism. Eventually, you’ll arrive somewhere.

It’s not optimism. It’s a description of what businesses without a destination actually do: they walk. Long enough that they accumulate real mileage, real spending, real effort. They build up an impressive body of activity. And they land, eventually, in a place they didn’t choose and couldn’t have predicted. Because they never decided where they were going.

That’s Marketing Drift in action.

The irony is that the destination itself is rarely the hard part. Businesses usually know, in broad terms, what success looks like. The harder thing is stating it plainly, committing to it in writing, and then holding every marketing decision up against it.

Before the next planning cycle starts—before the channel mix, the content calendar, the budget allocation—answer the Cat’s question.

“Where do you want to get to?”

The rest of the plan only makes sense once you have that answer.

Another all-time great, Yogi Berra, put it plainly: “If you don’t know where you are going, you might wind up somewhere else.”

And that, friends, is the real danger.

ChatGPT planned Mark Schaefer’s Paris trip down to the metro stop. It knew his gluten intolerance. It curated restaurant lists and optimized his itinerary for budget and taste. It did everything right.

He still flew Delta. Still stayed at his preferred hotels. Still packed his On Cloud sneakers. The AI did not get a vote on any of it.

That’s not a quirk. That’s brand doing exactly what brand is supposed to do. And it’s the starting point for everything Tom, Curtis, and this week’s guest unpack in Season 3, Episode 5 of Bullhorns & Bullseyes.

Mark is a marketing strategist, author of multiple books on brand, community, and AI, and the person behind the concept of content shock—which Tom spent six months thinking about before accidentally Googling it wrong. He’s also the reason Tom now subscribes to a newsletter he didn’t go looking for. That’s brand doing its job, once again.

The Sisyphean Task Nobody Wants to Hear About

Right now, a significant amount of marketing energy is being poured into one question: how do we get AI to mention our brand?

Mark’s answer is gentle but pointed. It’s a Sisyphean task—pushing the boulder up the mountain only to watch it roll back down. Most businesses don’t have the resources or the competitive positioning to win that fight. And there’s a more productive question available: what would make someone already know, trust, and choose you before they ever ask the AI anything?

Tom draws the parallel to the SEO era without hesitation. Companies spent enormous sums climbing search rankings, only to admit their best leads still came from referrals. Curtis, who spent the first decade of his career as an SEO specialist, doesn’t argue the point. The conversation isn’t anti-technology. It’s pro-strategy.

What People Buy Versus What You Sell

Mark has a habit of starting engagements by handing executives a piece of paper and asking them to finish a sentence: “Only we…” More than nine times out of ten, five people in the room produce five different answers. That’s not a messaging problem. That’s the absence of a marketing strategy.

The gap between what a company sells and what customers actually buy is where brand lives—and where most companies are losing ground. Mark tells the story of a regional telecom company whose team was obsessed with the sophistication of their technology. When he went out and talked to their customers, nobody mentioned the technology. What they said instead was: when something breaks at two in the morning, these people show up. They’re a security blanket.

The CEO was furious. Not at Mark—at himself, for spending five years unable to articulate what his company meant to customers, only to hear it solved in a single conversation. The exercise works because it forces you to go find the answer rather than assume it.

Curtis frames it as coherence before creativity. Until you know what you actually mean to the people buying from you, any creative or tactical work you do is speculative at best and wasteful at worst. That’s the chaos tax.

Why AI Can’t Do This Part for You

This is where the conversation gets direct. Curtis raises something Mark has written about: if you use AI for customer research, your competitors are running the same prompts and getting approximately the same output. The AI is reading the internet and producing a homogeneous answer. It’s doing its best job and that best job is, by design, consensus.

The 15 to 20 percent that actually creates competitive advantage requires talking to real customers. Not synthetic personas. Not AI-generated sentiment analysis. Actual conversations, conducted with some frequency, because what customers value this year may not be what they value next year. The market is not static.

Mark is careful to say this isn’t an argument against performance marketing, SEO, or advertising. It’s an argument that if those are the only tools in use, something foundational is being skipped.

The Emotional Continuum: From Ad to Community

Mark maps out a hierarchy that reframes how most brands think about their audiences. Advertising creates some emotional connection at the lowest level. Owned content channels—a podcast, a newsletter, a YouTube series—create an audience, which is a higher level of connection. Someone subscribing is raising their hand and saying they want more.

Community is different in kind, not just degree. The distinguishing feature: people in a community know each other. And here is the finding from Mark’s research for his book Belonging to the Brand that anchors the whole argument: when people form real relationships inside a brand community, that goodwill transfers to the brand. They belong to it. They do not leave.

The implication is that a company should invest as much in the connections between community members as it does in its own relationship with customers. That’s the architecture of advocacy.

Tom acknowledges the obvious counterpoint: none of this shows up cleanly in a report to the CFO. He and Curtis are clear-eyed about the challenge of making the case internally. Mark’s suggestion is to hand someone the book and let the business case make itself.

The Pandemic of Dull, and the Way Out

Tom introduces the concept of content shock—Mark’s term, from years ago, which Tom had mentally filed under “content tsunami” and spent months unable to find. The question he puts to Mark: in a world now flooded with AI-generated content, what does standing out actually require?

Mark’s answer is not “be human.” He calls that phrase hackneyed. His answer is closer to: approach the level of art.

Art, in this context, is not about craft or credentials. It’s about interpreting the human experience in a way AI cannot replicate. Mark shares the story of giving a five-hour lecture at McKinsey—his dream company, the pinnacle of what his career could produce—and feeling inexplicably depressed watching the sun set over Manhattan from the 64th floor of the World Trade Center afterward. He almost didn’t write about it. He was sure he’d be dismissed.

It became his biggest post of the year.

Not because he’s a great writer. Because the experience was real, specific, and recognizable. That is what AI is contributing nothing to. It is, in Mark’s phrase, accelerating a pandemic of dull. The content that rises above it is content that only a particular human, with a particular life, could have produced.

Curtis connects this to what he sees in his clients’ data. The posts that perform—an executive at a hospital bedside, a founder at their kid’s basketball game—are not performing because of the algorithm. They’re performing because they’re true.

Pesty, and What That Pest Control Company Did Right

Tom ends with a story about pest control.

A company knocked on his door, promised flexibility, and then wouldn’t let him out of his contract. He complained publicly. People came out of the woodwork with the same story. He swore off the category entirely.

Months later, a small ad appeared in a local newsletter. It said: stop letting your pest control company scam you.

Tom clicked immediately. He subscribed to the product—DIY pest control called Pesty—without reading a single performance claim or feature comparison. The packaging had puns. He is now an enthusiastic apostle, recommending it to everyone he can.

He never even checked whether it kills spiders more effectively than Ortho.

The company understood his anguish. They didn’t describe their product. They described him. That’s the whole lesson, compressed into one pest control ad. Mark calls it a ribbon on everything the episode covered. He’s right.

Watch and Listen to Season 3, Episode 5

Available on YouTube, Apple Podcasts, Spotify, and everywhere you listen.

Bullhorns & Bullseyes is hosted by Curtis Hays and Tom Nixon. Season 3 drops a new episode every two weeks.

The framework I borrowed and broke

I forced a Why once. It didn’t work, and it took me about a decade to understand why it didn’t.

When Simon Sinek’s “Start With Why” Ted talk came out, I watched it. The Golden Circle. Why-How-What. That’s his idea, not mine. And it’s a good one — which is exactly why we grabbed it. I was at an agency at the time, and we used Why-How-What as a culture thing. We sat down to declare our Why because it sounded right, because it was the kind of thing a serious agency was supposed to have.

We wrote one down. It read fine. It moved nothing.

I know now what that feeling is — that specific flatness when something is correctly assembled but somehow inert. At the time I didn’t have language for it. We’d done the work. We’d sat in the room, we’d talked it through, and what we produced looked exactly like what it was supposed to look like. Nobody called it wrong. But it never became the thing that shaped decisions. It sat in a document and aged out.

When I left, Why-How-What went with me — or rather, it went dormant. I didn’t think about it for years. I moved into consulting and carried no particular conviction about it one way or another. It was an idea I’d tried and shelved, the way you shelve a tool you bought for a job you couldn’t quite do.

Then Tom came back with it, and everything reordered.

has been doing brand work for clients for more than thirty years. He has the kind of patience with a problem that you can only earn by watching a lot of good-sounding ideas fail under real conditions. When he started working inside my client engagements, he kept coming back to two things. The first: “You are what your customer says you are.” The second, the one that I keep returning to: “I want to find out for myself, from the customer.”

That’s when the decade-dormant lesson broke open. Not because the framework was new — I’d carried it the whole time. It broke open because Tom wasn’t sitting down to declare anything. He was going to look.

He laid this out on tape during Bullhorns & Bullseyes Season 1 Episode 9 — “For All Intents & Purposes.” Thirty years of brand work distilled into one of the plainest truths I’ve heard in this industry. He said: “If it sounds to me like you’re looking in the mirror, turn that mirror into a window and look through it and see the customer or the user or the audience on the other side and talk to them about them.”

That’s the distinction. That’s the whole thing.

What I’d done at the agency was treat the Why like a mirror. Sit down, look inward, write what you see, call it true. What you get is a reflection of what you already assumed about yourself, dressed up well enough to sound like a discovery. The exercise feels like excavation but it’s only rearrangement. You’re moving your existing beliefs around the room and calling it finding something.

What Tom does is turn the mirror into a window. He doesn’t ask the brand what it believes. He goes to the people who buy and finds out what they already know. The Why was never inside us waiting to be phrased correctly. It was on the other side of the glass the whole time, in the customer’s own words, and the work was to go find it instead of invent it.

He puts the reason bluntly: “Few people care about you when they’re reading or consuming or viewing or deciding and shopping. They care about themselves.”

So the Why was never ours to declare. It belonged to them.

The same mistake, rebuilt in AI

This matters now because I’ve watched the same mistake reproduce itself in how people build with AI.

The default move is to write a posture into the system because it sounds right. “You are a helpful assistant.” You declare what the thing is supposed to be, the same way we declared what our agency was supposed to be, and for the same reason: it sounds good, and it’s what you’ve seen others do. Copied, not excavated. Invented, not found. It reads fine. It moves nothing.

I’ve been on the other side of that. When I built Duke — the orchestrator that runs my firm’s operating system — I didn’t get anywhere by declaring his character. The character came out of the actual decisions: the real client work, the corrections I gave him over and over, the way the firm actually behaves under pressure when something breaks. That’s what became the doctrine he holds. It came from the work itself, not the setup.

The Brand Guardian we build for a client works exactly the same way. A Brand Guardian is the document that governs everything the system says on the client’s behalf — it holds the positioning, the language, the reason people buy. It isn’t a posture we assigned or a vision statement we composed in a conference room. It’s the customer-validated reason people buy, excavated and written down so the system can hold to it. The research goes to the people on the other side of the glass. We listen to what they say about themselves — the language they use, the problem they’re naming, the outcome they’re actually after. And then we write it down as one thing.

AI did not generate it. AI revealed it.

There’s a version of this in search that I’ve watched play out in client work. Keyword research, done right, is not a creative exercise — you’re not naming what you think someone wants, you’re reading back the language of someone who already wants it and typed it into a box when no one was watching. The same move. Turn the mirror into a window. Go read what’s on the other side before you decide what to say.

Tom Nixon built his personal writing agent — Wradley — live, in about fifteen minutes, at a session we ran together recently. What he said afterward is what I want you to hear: “It took like 15 minutes, by the way... but it took months and months and months is the point — of you building this belief layer that made the writing part easy.” And then: “So now I’ve got an agent named Wradley who writes like me, but believes what I believe, thinks what I think... the thinking is baked into our doctrine, now into my writer. So when I ask it to write something, the thinking’s already there, all it’s doing is the typing.”

The fifteen minutes was real. So were the months. The months came first, and the months were not spent on the system. They were spent on the belief work that made the system possible. You cannot shortcut the months by getting better at the fifteen minutes. The sequence does not reverse.

Mira Bradshaw, a data tech and product leader, watching that same session sent something in live that is the sharpest version I’ve heard of why declaration fails:

“AI can give you the textbook theory, but it has never been burned. It has never been battle tested. You have. Give it those scars in context in your ICM system and see what happens. It’s game changing.”

Scars. Not principles. Not values. Scars. The things the system holds that only you could have put there, because only you were in the room when the lesson cost something. A declared posture has no scars. It has instructions. Instructions are other people’s scar tissue, averaged. They tell the system how to behave in situations someone else already anticipated. They do not give the system anything to protect when the situation has never been anticipated before.

There is something else from that session I want to use here. The night before, while I was working through a naming problem with my orchestrator, the system reached — unprompted — for the Fisher King parable. Pulled it from a B&B episode Tom had recorded two seasons earlier. Tom retold the parable in the session: “And the fool says, I didn’t know you were looking for a grail. I only knew that you were thirsty.” Tom’s word for what happened when the system surfaced it: indoctrinated.

Not retrieved. Not searched. Indoctrinated. The episode had been fed into the belief layer two seasons ago, and the system brought it forward when the moment called for it.

A declared posture cannot do that. A system told “you are a helpful assistant” does not reach for a two-season-old parable about thirst when the conversation calls for it. It has no way to know that thirst is the thing. A system that has been excavated and indoctrinated does know, because the parable was already inside it — not as a file to retrieve, but as a lens the system sees through. It did not know you were looking for a grail. It knew you were thirsty.

That is the difference between a posture and a worldview. A posture answers the question you asked. A worldview recognizes the question underneath it.

You cannot engineer a worldview worth holding until you’ve found the why that makes it true. The engineering comes second. The discovery comes first. And discovery is not a creative exercise. It is a research discipline. It is the decision to stop looking inward and go look outward, to treat the customer’s language as the source and your own assumptions as the thing you’re testing.

That sequence matters. Mirror without window — you get a system that performs confidence without having earned it. Window first, doctrine after — you get a system that holds to something real because the something real was found, not invented.

The treasure was already there

I gave away my entire OS map this week — the architecture, the agent roster, the folder structure, the belief layer outline. Posted it publicly. It cost me nothing, because the map was never the moat. I know that because I’ve watched independent builders with no access to my internal architecture keep arriving at the same shapes. Different people, different angles, same destination. When strangers reconstruct doctrine you already wrote without ever seeing it, the doctrine isn’t something you invented. It’s just how this actually works. Convergence is the proof it’s true.

Malcolm Guite — the poet-priest who just published Galahad and the Grail — puts it at the level of myth. He points out that almost every great quest story in the Western tradition is a story of acquisition: go get the golden fleece, bring back the fire, seize the treasure, return with power. That’s the template that runs from Jason and the Argonauts through most of what we still reach for. Then he says this about Tolkien: “He gives us an epic of letting go, of renunciation.” The treasure is there from page one. The whole journey is about not keeping it. That inversion isn’t a narrative trick. It’s the structure of the thing that’s actually true.

The Why works the same way. It isn’t acquired. It isn’t declared in a conference room. It’s already there — in you, and in how your customers actually behave. You go find where those two meet. We’re archaeologists, not architects. The belief is underground, not invented. The map just shows you where to dig.

Solve for Why

I spent years thinking a Why was something you sit down and write. It was the thing I copied, the thing I tried to invent, the thing I forgot about for almost a decade.

The day I stopped declaring it and started finding it was the day it got real.

That’s the whole reason this season of Bullhorns & Bullseyes is built around Solve for Why. Not declare it. Not phrase it well. Solve for it — the way you solve for anything you don’t yet have: by going to where the answer actually lives and bringing it back.

You did everything right.

You talked to customers. You sat through the interviews. You came back with something real… specific, emotional, true. And then it went into the machine.

Maybe it was a strategy session. Maybe it was a copywriter. Maybe it was an AI tool you fed the research into. And what came out the other side was: “We help leadership teams drive marketing accountability.”

Grammatically correct. Strategically neutered. Unrecognizable.

Curtis calls this signal loss. It’s not that anyone threw the research away. It’s that at every handoff—business owner to brief, brief to copywriter, copywriter to AI—a little more of what made it true gets smoothed away. Polished. Corporatized. Until it sounds professional. Which is another way of saying: it sounds like everyone else.

The Narrative Economy and You

Kristian Alomá came back for Season 3 to answer the harder question: once you have the truth, what do you actually do with it?

If you didn’t catch his Season 2 appearance: Kristian is a behavioral psychologist and brand strategist, founder of Threadline, and author of Start with the Story: Brand-Building in a Narrative Economy. The premise is that people are inherently storytelling creatures—we think in stories, dream in stories, and build our sense of identity through the stories we tell about ourselves. We don’t just buy products. We use brands to help shape that story.

Which means if your brand is the hero of its own story, there’s no room for the customer in the narrative arc. And if there’s no room for the customer, the relationship doesn’t go very far.

Most brands can’t stomach that inversion. It feels unnatural to talk about the customer’s story when you want the customer to know how great you are. But as Kristian put it: you’re either the date who talks about himself all night, or the one who asks questions and actually listens. One of those relationships goes somewhere.

The STORY Framework

Kristian walks through five elements every brand narrative needs:

Struggle. What problem are your customers living with before they ever heard of you?

Tool. Your approach or methodology—how you think about addressing that struggle.

Objective. What your organization is specifically doing to apply that tool to that struggle.

Reward. The measurable benefits they get.

Yearning. The emotional benefit. The glorious future. What life feels like when the problem is gone.

Most brands start at Tool and never make it to Yearning. They skip Struggle because it requires understanding the customer before talking about themselves. And they treat Yearning as soft—hard to put on a slide. But Yearning is where loyalty lives.

Nike doesn’t show you the shoe. It shows you someone who dreams of being great. Just do it. That’s Yearning. That’s why it works.

Where the Signal Dies

Two things go wrong, according to Kristian.

First: the insight lives in one room and nobody else hears it. Marketing has it. Sales doesn’t. The people delivering the service don’t. So the story that goes to market is disconnected from the story the customer actually experiences. The brand promise splits from the brand reality.

Second: insights don’t stick unless they’re translated into language each team can act on. What does this mean for sales? For the creative team? For product? A brilliant research summary filed in a deck nobody opens changes nothing.

What he calls the hymnal—a documented, codified source of truth that the whole organization can reference—is what keeps the signal alive. It becomes the creative brief. It’s the guardrail against drift. And in the AI world, it’s the context you feed into the system instead of just asking it to generate copy from nothing.

When you give AI the hymnal, it amplifies the right signal. When you don’t, it amplifies your mirror. Faster.

The Brands that Get it Write [sic]

Kristian walked through several real examples in the episode.

Nike makes everyday athletes feel like they belong to something. They’re not selling to professional athletes. They’re selling to the person shooting hoops in the driveway who wants to feel like Jordan. The gear becomes armor. The brand earns that feeling because it knows the Yearning—not the product specs.

FedEx understood that their brand lived with the drivers. Not in the warehouse, not in the logistics system…with the person who shows up at your door. So they invested in those drivers specifically.

Peloton’s 2019 “The Gift” ad missed the mark because it centered the brand—and a controlling-husband read on top of it—instead of centering the customer’s transformation. The closing line made Peloton the hero of the wife’s story. That’s the mistake.

McDonald’s, as Curtis put it, built something that made a mom with three kids in the back of a minivan feel like she had the situation handled. That’s the struggle. That’s the Yearning. Not the best ingredients. Happy meals. “We love to see you smile,” after all.

Are You Trying to Sell Me a Story?

Show me, don’t tell me. One of Tom’s favorite fallbacks.

You can tell me how great you are. How fast your product works. How “unparalleled” your service is. How you’ve got “the best people in the business.” (Right. Okay, sure.)

Those are all claims. Unverifiable to anyone who hasn’t already experienced them for themselves. That’s not marketing copy. That’s a naked sales pitch swaddled in braggadocio.

Or you can show me.

Show me the happier customer singing your praises…in their own words, with their own voice, through their own real, lived experience. Show me the hero who has already overcome the struggle that the next customer (read:hero) is starting to recognize in their own world. Show me the “unparalleled” commitment to service in real time. On video. Raw, and uncut…unscripted. Show me the struggle…the yearning…and, ultimately, the happily ever after.

That’s the difference between copy that’s trying to sell and stories you can authentically tell.

Because, put yourself in your own customer-shoes. You. How you shop. How you evaluate providers. Why you gravitate to one brand over another.

Is it the sales pitch? The shiny feature set? The bold, unprovable and wholly unverifiable performance claims?

Or is it the story that your favorite brands allow you to tell about yourself?

Watch and Listen to Season 3, Episode 4

Available on YouTube, Apple Podcasts, Spotify, and everywhere you listen.

Bullhorns & Bullseyes is hosted by Curtis Hays and Tom Nixon. Season 3 drops a new episode every two weeks.

Every conversation I’m in about AI advantage treats it like a tools problem.

Better prompts. Better models. Better workflows. Pick the right stack, wire it up right, and you pull ahead.

That’s the conversation. And it’s mostly wrong.

Your competitors have the same tools you do, and the same public data. NotebookLM, deep research, competitive analysis — none of it is yours, because all of it is theirs too. Anything every operator can buy or pull off the open internet can’t be the thing that sets you apart. When everyone can pull the same market intelligence in thirty seconds, the intelligence isn’t the moat. It’s the floor.

The moat is what they can’t get to. And the deepest part of it isn’t even the data. It’s the thing that produced the data.

Writing isn’t typing

My co-host Tom Nixon said something on the show this week that reframed the whole thing for me. We were talking about why generic AI content feels empty, and he stopped and made a distinction I haven’t been able to put down since.

Writing isn’t typing.

Writing is the thinking you do about the topic — for an hour, a week, ten years. That’s the writing. By the time you put it on pixels, that part’s just typing. Tom’s line was: keep the thinking to yourself, outsource the typing only.

Run that through the AI conversation and it turns the whole thing around. Most operators are racing to hand the AI more of the work. The whole pitch of the tools is that they’ll do the thinking and the typing both. Tom’s flipping it. The thinking is the only part that was ever yours. The typing was never the asset.

Which is why your conversations are the moat. Not because they’re transcripts nobody else has — though they are. Because they’re where your thinking already happened. The transcript is the residue of it. The decision you talked yourself into on a client call, the read you gave a deal in a debrief, the thing you finally said out loud in the last ten minutes of a two-hour session — that was the thinking. It already occurred. It’s sitting in the recording, waiting to be recovered.

What’s actually yours

There’s a category of data that is completely asymmetric. Nobody else was in the room. Nobody else was on that call. Nobody else heard what your client said when they stopped performing and started being honest.

Your meeting transcripts. Your sales call recordings. Your client debrief notes. Your team’s voice memos after a hard engagement. The conversations you have in your day that exist nowhere else and belong to no one else.

I was guilty of not capturing this for years. I sat across from clients and listened well. I made good notes. I carried the insight in my head into the next session and used it. The problem is that anything I carried in my head died when I moved on. The thinking happened — and then it evaporated. It didn’t compound. It didn’t train anything. It didn’t travel. It was experience that stayed private instead of becoming infrastructure.

That’s the first category of asymmetric data. There’s a second: the thinkers whose work you’ve spent years taking in.

The thinkers you’ve actually metabolized

Your worldview isn’t only what you think. It’s the synthesis of the people you’ve genuinely absorbed.

Tom named this on the show too. Our doctrine isn’t just our own thirty years of opinions — it’s Mark Schaefer, Brian Clark, Gini Dietrich. People we’ve interviewed, read, argued with in our heads, and adopted philosophically until their thinking became part of how we see. We stand on the shoulders of giants. The difference now is we can capture which giants, and how we metabolized them, instead of letting it stay vague.

That’s asymmetric too. A competitor can read the same Schaefer book you read. They cannot reconstruct which ideas you took, which you rejected, and what you did with them inside your own work over a decade. That synthesis is yours. It’s another layer of human data nobody can scrape, because the scraping already happened — in your head, on your terms.

So the raw material of your doctrine is two things. The conversations you’ve had. And the thinkers you’ve truly absorbed. Both are residue of thinking. Both belong only to you.

What fills the vacuum if you don’t

Tom put this next to something every brand person already knows. Your company is going to have a culture whether or not you take an active part in creating it, documenting it, policing it. Walk away from it and one forms anyway — just not the one you’d have chosen.

AI is the same. If you don’t form the doctrine, you don’t get a blank system. You get one that fills the gap from somewhere else — the model’s training data, the conventions baked into every borrowed template, the internet’s averaged opinion of what an operator like you probably believes. The thinking still gets done. It just gets done for you, by a machine running on everyone’s defaults instead of your convictions.

That’s the actual choice. Not “build the moat or don’t.” It’s “do your own thinking, or inherit the average’s.” The vacuum doesn’t stay empty.

Underneath the summary

When operators do capture transcripts, they extract the surface. They ask the AI to summarize what was discussed. They pull the action items. That’s useful. It’s also the shallowest possible return on the data.

There’s a question underneath the summary. Not “what did this person say?” but “what does this person’s pattern of decisions tell me about what they actually believe?”

I work with a home care operator named Mario. The surface read on his transcripts was clear — census, referral sources, staffing ratios. Standard operational concerns. But the pattern across a dozen conversations told a different story. Every time a decision carried reputational risk, Mario slowed down. Every time a decision was defensible, even if the economics were softer, he moved faster. He consistently chose the option that reduced his exposure to being blamed, even when the better economic option was on the table.

Mario doesn’t decide based on upside. He decides based on downside protection.

That’s not in the summary. That’s in the pattern. And when I understood it, everything changed — the framing of options, the order I put choices in front of him, which risks I named first. You only see that if you ask the data the right question, and you only have the data if you captured the thinking instead of letting it evaporate.

Where to start

If this feels like a lot, the first move is older than any of the tools, and you can do it today.

Keep one notebook.

Not a system. Not an app with folders and tags and a filing structure you’ll spend a weekend setting up and never touch again. One book — paper if you like the feel of it, a single digital note if you don’t. Everything goes in the next blank space: the thought in the car, the line a client said that you can’t shake, the read you got on a deal, the half-idea you’d normally lose by lunch. No judging it on the way in. No deciding whether it’s a good idea or a bad one before it lands on the page. The friction of “where does this go?” is what kills the habit — so it goes nowhere except the next open spot.

Two rules hold it together.

No to-do lists. The notebook is for thinking, not for getting things done. The minute you let tasks in, they crowd out the part that matters and the book becomes another inbox. Keep the chores on a scrap you throw away.

And review it constantly. This is the whole engine — not study, not memorization, just flipping back through it in line at the coffee shop, on the couch, between calls. That’s where it pays off. Ideas you wrote down weeks apart start to touch each other. A read you had on a client in March turns out to explain a pattern you spotted in June. The connections don’t live in any single entry. They surface when you keep passing back over the whole thing, and you only get to keep doing that because none of it evaporated.

This isn’t a productivity hack someone invented last year. It’s the commonplace book — the same practice behind Da Vinci’s six thousand pages, behind Darwin and Mark Twain. Da Vinci wrote a line at the start of one of his own that the artist Perry Daniels, who keeps the habit today and calls it the Grail Method, saved into his: “this is to be a collection without order, taken from many papers which I have copied here, hoping to arrange them later each in its place.” That’s the permission. It doesn’t have to be tidy to be the most valuable thing you keep.

One honest tension, because Daniels names it and I’m not going to dodge it: he’s against the structured second-brain — the Notion database, the elaborate filing system. And he’s right about the failure he’s pointing at. When structure becomes the place ideas go to die — sorted into folders so clean you never open them again — it kills exactly the connection-making the notebook is for. The structure earns its keep only when it does the opposite: when it helps you retrieve and collide what you captured, instead of burying it. The notebook is the raw, unsorted layer. The structure, if you add one, is just there to let you pass back over more of it than your hands could ever flip through.

Because that’s the move underneath all of this. Once you’ve got the corpus — the notebook, or the year of transcripts and voice memos behind it — you don’t ask it for a summary. A summary plays back what you already know. You ask it for the pattern across everything. The Mario read didn’t come from one call. It came from a dozen, asked the right question. Same one I ran on my own conversations before I ever pointed it at a client — I loaded my own calls and debriefs and made the AI mine me, and it handed back my worldview clearer than I’d have written it cold.

The Grail Diary

If you want to do that at scale, I built the thing I wished I’d had for years. It’s called the Grail Diary — the commonplace book, rebuilt so a machine keeps it with you. One notebook. Everything goes in the next blank space, no judging it at the door, no to-do lists allowed in. What the machine adds isn’t a verdict — it’s reach. It reads back over more of the corpus than your hands could ever flip through and lays the pieces out in front of you: here’s the line from March, here’s the one from June, here’s the read you gave that deal. It surfaces; you connect. It will not tell you what your pattern means, on purpose — the moment a machine names your belief for you, you’ve handed back the one part that was yours. Naming it is still your move. The mining I described above — pointing the AI at the whole pile and asking for the pattern, not the summary — runs on one discipline: when a pattern and a thing you said about yourself disagree, trust the pattern, and never let it hand you a belief you didn’t demonstrate. Empty beats fabricated, every time. I’m putting the Diary out in the open for anyone who wants it. Reach out to me if you’d like to try it. But you don’t need it to start. You need the notebook and the habit of reading back over it.

Public data is what everyone has. The conversations in your day, and the thinkers you’ve actually absorbed, are what nobody else can get to. That gap is the moat.

The thinking was always the asset. The only question is whether you’re keeping it — or letting the machine do it for you.

If this landed, two places to go deeper:

The conversations behind this thinking — [Bullhorns & Bullseyes]

The operators working through it live — Clief Notes

Join Tom and Curtis Friday at 10:00 AM EST, we host a live session for the real questions — the ones you don’t put in a DM. No deck. No agenda. 30–45 minutes of straight answers from two people who’ve spent years inside marketing and IT. — Join us at The Watering Hole

Not because it stops being useful. Because everyone will have it. The tutorials are everywhere. The frameworks are on GitHub. The syntax is not hard. You learn it the way you learned keyboard shortcuts — a little at a time, until one day you realize you cannot work without it and cannot remember when you acquired it.

That is not a criticism. It is a pattern. And I have watched this exact pattern run enough times now to know where it ends.

I remember when building a website was a real skill.

I spent fifteen years inside IT before I ever touched marketing — building networks, standing up servers, running the infrastructure that businesses quietly depended on and never thought about until it broke. Back then, if you wanted a website, you came to a shop like ours and we charged you real money for it, because building one took people who knew what they were doing. The skill was the moat. Then WordPress showed up, the CMS wave hit, and the thing we used to charge for became the thing anyone could do in an afternoon. The skill did not get less real. It just stopped being rare.

So I moved upstream. By 2016, before the big CRM platforms had attribution built in, I was working with a developer to stitch a closed-loop attribution system together out of cookies and Gravity Forms — because knowing what actually drove a sale was still rare, and the people who could see it were a step ahead of the people who could only build the form.

That has been the shape of the last fifteen years for me. Websites, then SEO, then social, then content. Every single time, the operators still talking about the tactic were already behind the operators who had moved to the layer underneath it.

Every time a capability becomes table stakes, the differentiator moves upstream.

AI is the same pattern wearing newer clothes. The prompt is the website. The skill everyone is rushing to learn right now is the skill that will be sitting in everyone’s hands inside three years.

The layer underneath prompt engineering is not better prompts.

It is what the system believes.

What a worldview actually is

I want to define the word before I use it the rest of the way, because worldview is one of those words people nod at while meaning different things.

A worldview is not an opinion. It is not a preference. It is not belief in the soft sense.

It is the instrument of vision. The frame of reference that determines which problems you recognize, which evidence you weight, which solutions you can even see. C.S. Lewis named it as cleanly as anyone has:

I believe in Christianity as I believe that the Sun has risen, not only because I see it, but because by it I see everything else.

That is worldview as instrument. Not the content of vision. The means of vision. The thing you see by, not the thing you see at.

Most people inherit a worldview without knowing they have one. The operators who actually shape their field are the ones who recognize they have one, name it, and install it on purpose.

That is what you are building when you build a governance layer correctly — the file that loads before the system does anything. Not a rulebook. A window the system sees the world through.

A rulebook is finite. The world is not. Every governance system built as a rulebook eventually meets a case the rulebook does not cover — and at that moment the system either defaults to its training (someone else’s worldview, averaged) or it freezes. A window does not have that problem. The window is not enumerating cases. It is determining what the system perceives in the first place. Cases the rulebook never anticipated still come through the same window and are interpreted by the same worldview. The system stays consistent because it is not consulting a list. It is seeing through a frame.

That is worldview engineering. And it is the discipline almost nobody is working on yet, because almost everybody is still upstream of it — still optimizing models, tuning retrieval, wiring orchestrators, arguing about which frontier model runs fastest.

All of that is real. All of that matters. None of it is the moat you think it is.

What happens when the tools converge

Picture two operators in the same market. Same industry. Same customer size. Both are at Level 4 on the AI ladder. Both have orchestrators. Both have specialist agents. Both have persistent memory and clean file systems and retrieval that works.

They buy access to the same frontier model. They borrow the same folder architecture from the same GitHub repo. They watch the same tutorial. They run the same setup.

Six months later, one of them is producing work that feels unmistakably like a point of view. The other is producing work that is technically flawless and completely forgettable.

Same piano. Different music.

The difference is not in the instrument. The instrument is identical. The difference is in what the system was told to believe before the first prompt was ever written.

I have shown this to a room. Same model, same prompt, run twice — once with a blank system, once with the belief layer loaded. Two completely different outputs sitting side by side, and the only thing that changed between them is what the system believed.

One operator did the belief work. The other downloaded the scaffold and assumed that was the same thing.

It is not.

Infrastructure can be copied in an afternoon. A worldview takes years to earn and can only be written by the person who lived it.

The invisible layer most systems are missing

If you open the CLAUDE.md file or the system prompt or whatever governance layer sits at the top of most advanced AI stacks, you will find something like this:

You are a helpful assistant. Be concise. Maintain professional tone. Follow these SOPs.

That is a posture. It is not a worldview.

Here is what a worldview-engineered system has that a posture-governed system does not.

It has opinions about what quality means. Not “produce good work” — a specific, defended definition of good that would let the system reject something most operators would approve.

It has a list of things it will never say. Not because they are technically wrong. Because they are inconsistent with what the system believes about its customer and its work.

It has a hierarchy of tradeoffs. When speed conflicts with depth, what wins? When the client asks for something that contradicts the doctrine, what happens? When the data suggests one thing and the belief layer suggests another, which gets the next question?

It has an emotional register. What tone does this system protect? Not “professional.” What does that actually mean for this operator, for this customer, in this moment?

It has a definition of originality. What would make this output something nobody else could have produced?

Most systems have none of this. They have instructions. Instructions tell the system how to behave. A worldview tells the system what to protect when the instructions run out.

And the instructions always run out.

Why you cannot download someone else’s worldview

This is the thing I have to say plainly, because the temptation in every builder community is to find the best stack and replicate it.

When you download someone else’s CLAUDE.md — their role files, their doctrine layer, their agent prompts — you are not just importing their architecture.

You are importing their beliefs.

Every structural decision in a governance file encodes a conviction. What gets preserved and what gets filtered. What the system treats as signal and what it treats as noise. What “good” means. What “done” means. What the system is never allowed to optimize at the expense of the business.

Those decisions were made by someone who lived something. Who earned a specific set of rules by watching a specific set of things fail. The rules look like technical choices. They are not. They are formed convictions dressed in syntax.

If their convictions happen to match yours, you got lucky. If they do not, the system drifts toward their answers every time you do not have an explicit override. And you will not always know which decisions were value-laden and which were just engineering.

The danger is not that borrowed scaffolding breaks. The danger is that it works well enough that you never notice the beliefs underneath are not yours.

You end up with a coherent system. Coherent around someone else’s doctrine.

The output is competent. The output is on-time. The output also has no gravity. No invisible weight that makes a reader feel like only one person in the world could have written it.

That gravity is a worldview. It cannot be forked from a repo.

What the belief work actually looks like

I did not sit down one afternoon and write a worldview. Nobody does.

I have a podcast. Seventy-five episodes across three seasons. Hundreds of client conversations. A decade of diagnoses that felt like judgment until I stopped and named what was underneath them.

The doctrine was already there. AI did not generate it. AI revealed it. I dumped the transcripts and asked the system to find the patterns. What I was looking for was content ideas. What I found was doctrine. Rules I had been enforcing for years without naming them. Beliefs I had been protecting without writing them down.

I had been the last one to hear them.

Then I wrote them down. Not a brand guide. Not a slogan. The actual rules. Why diagnosis comes before tactics. Why no client gets a creative brief before the customer reality is clear. Why every output has to be traceable back to something true about the customer. Why certain things are never said, regardless of whether they would convert.

Then I encoded them. Not into a prompt. Into a governance layer the entire system loads before producing anything. Every agent. Every session. Every output.

The system changed when those files went in. Not because the model changed. Not because the architecture changed. Because the system finally had something to be faithful to that was mine — not the internet’s average.

That is the sequence. The belief work comes first. Always. Before the folder system. Before the orchestrator. Before the first agent is named.

Most people do it in the opposite order.

They build the system and then try to pour belief into it later. But the system was already shaped by defaults — the model’s training data, other people’s scaffolding, the conventions baked into every template they borrowed. The architecture has already made a hundred quiet decisions about what matters. By the time the operator tries to install a worldview, they are fighting the gravity of a system that was built without one.

The chest has to be built before the brain is turned on. Otherwise you get intelligence without conviction.

You are designing epistemologies now

There is a word for what happens at the far end of this discipline.

The really advanced builders — the ones two or three moves ahead of the current conversation — are not designing workflows. They are designing epistemologies.

They are deciding how truth is evaluated inside a system. What counts as evidence. How ambiguity is handled. What signals deserve trust and which ones get questioned. What emotional outcomes matter beyond task completion. What identity the system is protecting on behalf of the operator who built it.

That is not hyperbole. That is the literal description of what a governance layer does when it is built from a genuine worldview rather than borrowed instructions.

And it compounds.

A system with a clear worldview gets more itself over time. Every output that holds the doctrine reinforces the doctrine. Every decision the system makes that is consistent with the operator’s beliefs adds weight to the next decision. The outputs start to feel related even when the topics change. There is an invisible gravity.

That is what you feel when you read someone who has a coherent perspective. Every piece carries the same weight. You do not have to read the byline. You already know who wrote it.

AI systems will work the same way.

The companies that win the next phase of this will not be the ones with the most capable models. Capability will be ambient, like electricity. You will assume it is there and pay a flat rate for access.

The companies that win will be the ones whose systems consistently express a recognizable philosophy.

Not a brand voice. A philosophy. A set of beliefs about the customer, the work, and what the work is for that is so clearly encoded into the system that every output carries it — whether a human or an agent produced it.

Humans are still the source

I want to name one thing before the close, because it is the part that gets lost when people hear “worldview engineering” and imagine they are being asked to turn AI into something it is not.

AI can scale judgment. It cannot originate lived conviction.

That is not a limitation the models will eventually overcome. That is a structural fact about where belief comes from. Belief comes from consequence. From watching something fail. From the client who taught you the lesson you did not want to learn. From the year the business nearly didn’t make it. From the thing you almost said and stopped yourself. From the rule you earned by breaking it.

AI has no biography. It has training data. Training data is other people’s consequence, averaged. It is powerful. It is not yours.

This is why generic AI content feels empty. It has syntax without sacrifice. It has structure without stakes. There is no accumulated life inside it. The system has no skin in the game because it has no game. It has only examples of games other people played.

Mark Schaefer built a version of himself you can talk to — a ChatGPT he calls Markbot. The thing that struck me was not that it knew more than any other model. It did not. It had read the same internet everyone else read. What it had was a soul. It carried his worldview so completely that asking it a question felt like asking Mark a question and getting Mark’s answer back. The information was ambient. The worldview was the whole point.

The operators who understand this will stop asking how to make AI sound more human. They will start asking how to make AI more faithfully theirs.

That is a different question. It leads to a different discipline.

It leads here.

and I have been working out this argument for months. He has spent thirty years inside brand strategy. I have spent fifteen years inside IT systems and fifteen inside digital marketing. We came at this from opposite directions and arrived at the same place.

He chases truth in meaning. I chase truth in data. The worldview lives where those two lines cross.

He said something to me early on that I have not been able to improve on.

I don’t care what you think you are. You are what your customer says you are.

That is the discipline I am describing in commercial language. A worldview is not what you believe about yourself. It is what you have heard from the customer often enough to stop arguing with, and have written down with enough precision that a machine can defend it when you are not in the room.

He also said this, later in the same conversation:

The doctrine without a machine to amplify it is useless. A machine that amplifies without a doctrine to guide it is also useless.

That is the whole thing. Two sentences. The first three pieces in this series climbed the ladder. This piece names what the ladder is leaning against.

What Jobs saw in 1985

There is one more piece of this I want to put on the table before the close.

Steve Jobs gave a talk at Lund University in 1985. The Mac had shipped. He was working on what would become NeXT. He was talking to students about computers, and he stopped to talk about Aristotle.

He said that as a kid, books were what kept him out of trouble. He could read what Aristotle wrote without an intermediary. The book was a miracle. The source got to the destination with nothing in the middle. But he could not ask Aristotle a question. He named it as the limit he wanted the next generation of machines to break.

He used the exact word:

My hope is someday, when the next Aristotle is alive, we can capture the underlying world view of that Aristotle in a computer, and someday some student will be able to not only read the words Aristotle wrote, but ask Aristotle a question and get an answer.

He was not predicting LLMs. He was naming the prize of the computer age four decades before the machines existed that could deliver it. The prize is the encoding of a worldview precisely enough that it can be queried after the operator is gone.

The words are the artifact. The worldview is the thing that made the words mean something. For two thousand years we have had the words. We have never had the worldview. Aristotle’s actual frame of reference, the instrument he used to see the world by, survived only as far as his apprentices could carry it, and then it filtered through centuries of translators, summaries, and arguments about what he probably meant.

Jobs saw, in 1985, that the computer might finally close that gap.

Forty years later, we are the operators standing in front of it. The technology exists. The discipline does not yet. The companies that build the discipline are the ones who will encode worldviews precisely enough that the systems can run on them after the founder is in another meeting, or another company, or no longer in the room at all.

That is not productivity. That is not workflow automation. That is the move Jobs named four decades before the machines existed that could actually do it.

This is what worldview engineering is for.

The question that decides which operators win

Prompt engineering is how you use the tool.

Worldview engineering is how you decide what the tool is for.

One is a skill. The other is a discipline.

The skill will be table stakes. The discipline will be the moat.

The question in front of you is not which model to use or which agent framework to build on.

The question is whether you have encoded the worldview that makes you know which problems are worth solving in the first place.

And whether you wrote it down before you turned the machine on.

Get more from Tom Nixon in the Substack app

Available for iOS and Android

Get the app

Something is off and you cannot name it.

Your output is faster than it has ever been. Your folders are organized. Your agents have names. The work that used to take a day takes an hour. And quietly, in a place you do not talk about on LinkedIn, you suspect it is still generic.

It reads competent. It reads on-brand. It also reads like it could have come from any other operator in your space who built the same folder system this quarter. You thought the architecture was the answer. The architecture is built. The output is still drift.

You do not know what is missing. You do not know what to ask for. You are not behind. You are standing in front of a piano.

The piano in the house

A lot of people had a piano in the house growing up. Very few of them could play it.

We had one too. It sat in the front room of my parents’ house for years. The instrument was there. The bench was there. The lid opened. The keys worked. None of that made anyone in my family a musician.

I was a senior in high school the night I figured out what a piano was actually for. A few of us had come back to my friend Scott’s house after closing down the pizza place where one of the guys worked. We had custom pizzas and cans of Coke and nowhere to be. Scott sat down at his family’s piano and started playing the theme song to Cheers. None of us knew he could play. A roomful of teenage boys on a Saturday night, singing where everybody knows your name. He never played for us again. I have never forgotten it.

The instrument had been in his house his whole life. The instrument was not the music. Scott was the music. The piano was just what let him give it to us.

This is the part of the AI conversation almost nobody is having.

The folder system is the piano. The orchestrator, the specialists, the file routing, the persistent memory. All of it is the instrument. It is real. It is capable. It is sitting in your office right now if you got to Level 4.

The score is what you play on it. The score is the doctrine. The score is the set of beliefs and operating rules your agents are protecting every time they produce work. Without the score, the instrument is furniture. Beautiful furniture. Expensive furniture. Furniture all the same.

The ICM is the instrument. Doctrine is the score.

That is the line the rest of this piece pivots on. I want you to feel it before I name it again.

Why your SOPs are not the score

Here is where most of the operators reading this are going to nod and close the tab. I already do this. I have my SOPs loaded into the project.

You do not. And the distinction is the whole conversation.

Picture two agents sitting next to each other. Same model. Same architecture. Same access to the same CRM. The prospect on the screen is a name you have been chasing for three weeks. The qualification call just ended. The agent’s job is to draft the follow-up.

The first agent has your SOPs loaded. Subject line formulas. Paragraph cadence. The opener you like. The signature you use. The constraint that says no em dashes and no exclamation points. The agent writes a clean follow-up in your voice. It is on-brand. It is the right length. It would pass any per-asset review you put it through.

The second agent has your SOPs loaded too. And underneath those SOPs, it has your doctrine. The rule that says we do not follow up on prospects who failed the qualification frame. The rule that says we do not deepen relationships that violate the ICP, no matter how warm the conversation felt. The rule that says a beautifully written wrong move is more expensive than a missed one, because it commits us to a customer the system was built to refuse.

The second agent reads the call notes and stops. It does not draft. It surfaces a flag instead. This one failed the frame. The follow-up the SOP is about to fire would deepen the wrong relationship. Confirm before we proceed.

The first agent is skilled. The second agent is right.

SOPs tell the system how to do the work. Doctrine tells the system what the work is for, and which work is not worth doing at all.

Most operators stopped at the first agent and called the job done. The output looked good. The output kept coming. They never noticed the system had no opinion about whether the output should exist. SOPs make the agents skilled. Doctrine makes them right.

What took me a decade to see

I did not know I had been building it.

For thirteen years I sat inside paid media campaigns. I watched ad accounts and CRMs and the gap between what marketing promised and what sales actually closed. I wrote diagnoses. I made calls about what to fix first.

I don’t remember the client. I remember the feeling. I was walking through another account, another CRM gap, another campaign that looked busy but wasn’t connected to revenue. Halfway through the diagnosis I realized I wasn’t deciding anymore. I was following an order.

The client had come to me for paid media. The one before had come for SEO. The one before that, the same. I had been taking the order and getting to work. I had stopped asking whether the order was the actual problem.

So I stopped. I started asking a different question instead. Do you know why your customer buys from you? And almost every time, the answer was no. Not a sharp no. The kind of no that sits in the air for a second while the person on the other side of the desk realizes they have been running a business for fifteen years on an answer they never wrote down.

That question became a rule. I do not take the order at face value anymore. Not paid media. Not SEO. Not the brief they walked in with. The brief is almost never the problem. The problem is upstream of the brief, and the operator usually does not know it yet.

It had been a rule the whole time. I had been treating it as judgment. It was doctrine. The repetition had been the proof. I had been the last one to notice.

That was when I started writing things down. Not framework names. Not slogans. The actual rules. Why diagnosis comes before tactics. Why no client gets a creative brief before the customer reality, measurement layer, and strategic constraint are clear. Why nothing leaves the building without governance. Why revenue has to be the number that eventually proves whether the work was real.

I never sat down to write a doctrine. I had been writing it for years and did not know. Seventy-five episodes of a podcast with my co-host Tom Nixon, talking with operators and clients and peers about the work. The beliefs spilled out in conversation. The vulnerabilities spilled out in conversation. The rules I had earned by repetition spilled out in conversation. None of it was organized. None of it was named.

Then I started mining the transcripts.

I dumped the episodes into AI and asked it to find the patterns. What I was looking for was content ideas. What I found was doctrine. The same rules I had been saying out loud to a microphone for three years. I had been the last one to hear them.

That is the part of this story I want you to sit with for a second. The doctrine was already there. AI did not generate it. AI revealed it. The amplifier the manifesto named two essays ago is not just an output amplifier. It is a recognition amplifier. It made me visible to myself.

Then I found the folder system.

I watched what people like Jake Van Clief were doing with persistent context and agent architecture, and the air left the room. I had spent a decade preparing for the moment I would have somewhere to put it.

Most operators are climbing the ladder in the opposite order. They are installing the instrument first. They are watching demos of other people’s pianos and buying the same model. The instrument arrives. They sit down. They realize they do not have anything to play.

The order matters. Context precedes architecture. The manifesto pointed at this. The 5 Levels mapped the climb. This piece names what you climb toward: a doctrine strong enough to become infrastructure.

The rules a system has to honor

There are three lines I keep coming back to. They are not slogans. They are the rules a system has to honor before it produces work worth shipping.

Coherence before creativity.

If your underlying logic does not line up, AI does not fix it. AI runs the incoherence at higher volume with better grammar. The manifesto named this. It is still true at Level 4. It will still be true at Level 10.

Diagnosis before tactics.

The glitch in a marketing system is almost never where the operator thinks it is. It is upstream. It is in how the customer is defined, or how the data flows, or how the story is told, or how the close-loop is wired. An agent told to write a better ad without a doctrine that demands diagnosis first will write a prettier version of the wrong thing.

Context before architecture.

This is the one this piece exists to name. The folder system is downstream. What lives inside the folders is upstream. If the upstream is empty, the downstream produces drift. If the upstream is sharp, the downstream produces leverage.

Three rules. Two of them earned the second appearance. The third earned the first.

Doctrine is not a brand guide. It is not a folder of examples. It is the operating law underneath the work. Who we serve. What we refuse. What must be true before we act. What proof matters. What the system is never allowed to optimize at the expense of the business.

What changes when the score shows up

There is an operator I am working with right now. Services business. Mid-market. Twenty plus years in. He read the manifesto. He read the 5 Levels. He showed up at Level 3 already.

He had agents. He had a folder system. He had done the climbing. He thought he had a marketing problem.

We ran the diagnostic. He did not have a marketing problem.

He had a sales problem.

The way he sold to a prospect was nothing like the way his team sold to the same prospect a week later. He sat down with people and listened. He took a pain inventory. He used no acronyms and no feature lists. He helped the prospect see the shape of the problem before he ever mentioned a service. By the end of a first meeting, the prospect felt understood in a way three other firms had not managed.

Then the handoff. His team would walk in with a feature list and a thirty-six-page proposal. The prospect who had just heard the owner help them think about their business would now be asked to read a document they could not read. My partner Tom Nixon named the pattern early. Selling simplicity, delivering complexity. The brand promise was being broken in the second meeting, every single time.

Deals were falling apart at the handoff. He did not know why until the diagnostic surfaced it.

He had been protecting the way he sold for twenty years. He had never written it down. None of it was protected. The agents he had stood up could not hold a register he had never named.

Now we are building the doctrine. Not the SOP for the follow-up email. The doctrine underneath the sales motion itself. What he listens for in the first ten minutes. The frame he uses to make the prospect’s situation make sense to them before he ever opens a deck. The reason he refuses to send a proposal in the first meeting. The reasons he walks away from deals his team would chase.

When the doctrine is written, his sales team will sell the way he sells. The agents will protect the way he sells. The handoff will stop bleeding.

Same instrument. Different output. The agents will not have changed. The score will have.

The architecture you have already seen

In my system, there is an orchestrator. There are specialists with defined roles, constraints, and judgment boundaries. There is a doctrine layer the team loads before producing. We call it the Brand Guardian. It is the operating law every specialist is held against. There is a verification step the producer is never allowed to close on its own work. There is a memory layer that stays human-readable so the next operator can open a file and understand exactly what was being protected.

The names matter less than the order. Mine has an orchestrator named Duke who routes the work before the work begins. The names are mine. The pattern is portable.

A note for the builders, because this is the part I have to clarify almost every time it comes up. The doctrine layer my team uses is one thing. The MCP I built is something else. The MCP is how I deliver doctrine into a client’s AI environment so their internal teams can produce work that holds the brand without my team in the room. Same doctrine. Different surface. The reason both live behind controlled layers is the same reason a coach does not hand his playbook to the opposing team. The doctrine is the asset. If I give it away, I give away the system that produced it.

The score is enforced at the door, not after the fact. The system assumes the producer is the worst auditor of the produced. The memory lives in folders, not in a vector database, so the doctrine stays readable to a human a year from now.

That is the instrument. None of it matters without the score.

What you actually climb toward

Most of the AI ladder discourse is about building more. More agents. More automations. More overnight workflows.

That is not the path. Or it is, but only after the harder path.

The harder path is naming what you have been protecting all along. Most operators have a doctrine in their head. They built it by repetition. They have never written it down. They are about to install an instrument and discover they do not know what to play.

Do not do that.

Most operators do not need more agents yet.

They need a score.

They need to name the rules they have already earned.

Then, and only then, they should build the system that protects them.

The strongest AI systems in the next few years will not belong to the operators with the fanciest orchestrators. They will belong to the operators with the clearest philosophy embedded inside them. Mark Schaefer has been saying a version of this for years. Technology amplifies what already exists. It does not replace the need for humanity, originality, or a point of view.

AI systems without doctrine will produce competent mediocrity at scale.

AI systems with doctrine will produce something nobody else can.

Coherence before creativity.

Diagnosis before tactics.

Context before architecture.

That is the system worth amplifying.

So I want to leave you with the question I had to ask myself before I knew there was a question.

If I cloned your stack tomorrow. Same agents. Same prompts. Same file structure. What would still make your work unmistakably yours?

If the answer is nothing, the work in front of you is not building more agents.

The work in front of you is naming the doctrine you have already earned.

Then go build the system that protects it.

The most dangerous sentence in business is not, We don’t know need to know our customer.

It’s: We already do.

That’s what I said at the top of this episode, and I meant it. Because the second sentence closes the door. The first sentence at least leaves it open.

Emily Bielak came back to Bullhorns and Bullseyes this season to talk about what happens when you actually go find out what you don’t know — and, more importantly, what to do with it when you have it.

Research that Sits in a Filing Cabinet Tells No One

Curtis started the episode with a business owner he’d been sitting across from recently. The owner had a two-step process to acquire customers. Step one: qualify them. Step two: convert them. The qualification rate was improving. The conversion rate wasn’t. And the owner was deep into pricing strategy and discount structures trying to fix it.

Curtis’s question: When did you last talk to a customer? Not survey them. Talk to them.

Emily’s answer to why that question gets resisted so often: emotions are perceived as fluffy. Okay, we know how they feel. What does that mean for revenue?

So the Martec Group built a way to quantify it: The Martec Emotion Score.

It works like NPS: measure the net level of pleasantness of emotion, track it across segments, across journey stages, compare it to competitors. Now it’s a metric, not a feeling. Now the CFO can engage with it.

But the number is only half of it. The other half is the so what and the now what.

Emily laid out the framework: What, So What?, Now What?.

Step one is what you heard: 60% of people feel happy at this stage.

Step two is what it means: So, what’s the implication?

And step three is what you do: Now what? How do you move that number to 75%?

That last step is where most research dies. Someone hands the client a data dump, calls it done, and moves on. The research sits. Nothing changes.

Segments, not Monoliths

One of the things Emily pushed on hardest was the tendency to treat all customers as one thing. Leaders who say “we know our customer” are usually thinking of one customer—their ideal customer…the 10 % of their customer base that fits the perfect profile. And they’re thinking of that customer at the point of purchase, which is the end of the journey—a destination, not the journey itself.

Real segmentation surfaces complexity that a single ideal client persona (ICP) misses. Emily gave an example from a barbecue brand they worked with. The ICP was less than 10% of the population they studied. The other 90%—some buying only sauces, not seasonings, some in adjacent segments that could be captured with the right product or message—were invisible.

The kicker: different segments going through the exact same process can have completely different emotional journeys. Being put on a wait list might be deeply frustrating for one segment and completely neutral for another. You can’t design for that if you’re only thinking about your ideal customer.

Peak-End Theory

Tom brought up Peak-End Theory—the principle that how people remember an experience boils down to how they felt at its peak moment and how they felt at the end.

Emily’s example from her own life: Machu Picchu. The peak was walking through the gates. The end was a 12-hour flight delay sleeping in the airport. Years later, those are still the two things she remembers: the extraordinary high and the miserable close.

Apply that to a business. A prospect comes in at peak anxiety. The neck is bleeding, they need help, the business owner is brilliant at meeting them there, calming the room, and handing them off. Then the sales process takes over. It introduces complexity. Features. Layers of information. And the close rate is terrible.

Tom named it directly: the peak was calm and reassurance. The end was complexity and fear of the unknown. Those two things don’t match. The research shows you where the journey breaks. The execution is just fixing the break.

The OODA Loop

Curtis brought in a framework he uses from military aviation: Observe, Orient, Decide, Act. Developed by a Top Gun instructor for fighter pilots making split-second decisions in combat. Observe the data, orient yourself around what it means, decide what to do, act. And then loop back.

The part Curtis emphasized was the act. Someone has to be accountable for the decision that gets made. And the loop means you come back to the numbers to see what actually happened after the decision was implemented. Research without a loop is just a snapshot.

How often do you need to do this?

Curtis asked the practical question: with everything changing this fast, how often does a company need to revisit customer research?

Emily’s answer: minimum viable is customer interviews every year—five, ten, or more depending on your customer base. Just someone who isn’t the main point of contact asking real questions about what’s going on. It’s free to do and almost nobody does it.

The fuller version—a qual>>quant>>qual segmentation study—Emily would recommend you revisit every three to five years, or sooner, if the market gets disrupted. Qual interviews to surface context. A statistically significant survey, at least 800 respondents for segmentation. Then qual again to bring the segments to life.

She made one other point that landed. For the business owner who says they can’t afford this: If you’re spending $120,000 a year in marketing and 50% of it is misaligned with what actually drives your customers’ decisions, that’s $60,000 wasted. The research isn’t a cost. It’s an investment that makes the spend worth deploying.

Check Out Season 3, Episode 3

Available on YouTube, Apple Podcasts, Spotify, and everywhere you listen:

https://bullhornsandbullseyes.substack.com/podcast

Bullhorns & Bullseyes is hosted by Curtis Hays and Tom Nixon. Season 3 drops a new episode every two weeks.

You’ve been using ChatGPT for six months. Maybe a year. You’re getting real work out of it. And somewhere underneath that, quietly, you suspect you’re not getting what everyone else on LinkedIn seems to be getting.

You watch other founders post their stack. Their workflows. Their overnight agents. You read it and something doesn’t add up against what you’re actually doing at your keyboard, which is mostly typing into a chat window and hoping the next answer is better than the last one.

You’re not behind. You’re stuck. And there’s a reason for it that nobody is saying out loud.

The reason is that there’s a ladder. Five rungs. And almost everyone using AI right now is standing on the first one without knowing the other four exist. The hype on your feed is being written by people on rung three and four describing their view, while you’re on rung one wondering why the view doesn’t match.

I wrote a few months back that AI is your amplifier, not your prophet. That if your underlying system is coherent, AI compounds growth, and if it isn’t, AI compounds waste at 100x speed with better grammar. I stand by that. What I didn’t say is where you actually start climbing, and what each rung up the ladder actually looks like.

That’s what this is. Five levels of AI mastery. Most people are stuck on the first one. They don’t know it. And it’s costing them every month they stay there.

Level 1. The Intern With Amnesia

This is where almost everyone is right now.

You open ChatGPT. You ask a question. You get something useful. You ask another. After ten or twenty back-and-forths, you’ve gotten somewhere real. You feel like you finally have an AI that gets you.

Then you close the tab. Or you come back the next morning and start a new chat. And everything you built is gone.

So you do the thing I keep watching people do. You stay inside one chat for a year. Hundreds of exchanges. Scrolling for context. Terrified to close it because the memory you built lives there and nowhere else.

I talked to someone last week who’s been working in the same chat for thirteen months. Building a report. Hundreds of messages deep. She’s not afraid of AI. She’s afraid of losing what she taught it.

That’s Level 1. Think of it like an intern.

Brilliant. Speaks twelve languages. Can code. Solves problems faster than any human you’ve hired. You have a great first day with this intern. You walk them through the company, the clients, what good looks like. You feel productive.

The intern goes home. Comes back the next morning. Forgot everything.

Same skills. Same intelligence. Zero memory of you.

That’s the problem with basic chat. The AI is passive. It waits for you to ask. And every time you start a new conversation, you’re training the intern from scratch.

You can learn prompting here. You should. But prompting alone won’t save you. Because no matter how good your prompt is, you’re going to type some version of it again tomorrow.

Level 2. The Assistant That Remembers

This is the most important shift in the ladder. If you do nothing else after reading this, do this one.

Inside ChatGPT and Claude there’s a feature called projects. Gemini calls them gems. They look like folders. They’re not folders.

A project lets you write your prompt once. Define the role. State the objective. Load the context. Add source files. Brand voice rules. Past work. Examples of what good looks like. Then every chat you start inside that project loads all of it automatically.

The intern stops having amnesia.

Now when you talk to your AI, you don’t open with a 400-word prompt. You open with natural language. “Let’s research this company.” “Draft the follow-up.” “Pull the patterns from these three calls.” The context is already loaded. The persona is already built. The constraints are already in place.

Here is what that looks like for me. I have a project called Paula. Her job is prompt engineering. Loaded into Paula are the prompting frameworks I trust, examples of strong prompts, and the constraints I want her to honor. When I open a chat with Paula, I do not type a setup paragraph. I type one line. “Paula, build me a prompt for audience research for a marketing agency that also does revenue ops. We will use it with ChatGPT deep research.” Paula comes back with a structured prompt that runs to the right length, in the right format, against the right model, every time. Base ChatGPT will give you a serviceable answer to that same request. Paula gives me a prompt I can actually ship.

This is also where you start loading expertise you trust into your assistants. We recently had Will Leach from MindState Group on Bullhorns & Bullseyes. His whole framework is about the psychological mindstates that actually drive customer behavior. The kind of thinking that takes years to internalize. Loaded into a project, that framework becomes context your AI applies every time you ask it about a customer. You stop getting generic. You start getting Will. (We sat down on the podcast. Link in the comments.)

The shift from Level 1 to Level 2 is the shift from interviewing a new intern every day to having an assistant who actually knows you.

Most people skip this and try to jump straight to the fancy stuff. Don’t. Spend real time here.

Level 3. The Coworker With File Access

Level 3 is when AI stops living inside a browser tab.

Claude Desktop and a few similar tools opened this up. You point the AI at a folder on your computer. Now it can read your files. Write to them. Modify spreadsheets. Update Word documents. Reorganize a messy downloads folder. Create new files where it needs to.

You stop uploading. You stop downloading. You stop dragging files in and copy-pasting outputs back. The AI is working alongside you, not across a glass wall from you.

Here is what that means in a real workday. I got off a call with a client a few weeks ago. Hour and a half. They wanted to pivot in a couple of directions. Three different deals coming out of one conversation. Normally that turns into an hour of cleanup. Open HubSpot. Create the deals one by one. Type the notes. Build the follow-up tasks. Try to remember everything they said.

I did none of it. I told my AI what happened on the call. It went into HubSpot, created the three deals, set the to-dos, and filed the call notes. I reviewed it and moved on.

That is the Level 3 jump. Not faster typing. Not a better prompt. The friction of moving information between you and your tools collapses, because the AI is allowed to touch the tools.

This is the orange belt. Useful. Real. Most people who get here feel a meaningful jump in productivity. But it’s still one-to-one. You and the AI. One task at a time.

The bigger shift is next.

Level 4. The Orchestrator

I talked to a builder recently who runs an AI platform serving billion-dollar companies. He told me he has thirty-one AI employees built. He doesn’t use most of them.

That is the wall before Level 4. You can build a stable of agents and still be the bottleneck, because you are the one routing every request by hand. Project to project. Tab to tab. Trying to remember which assistant knows which thing. The agents are real. The orchestration isn’t.

Level 4 is when you stop talking to a single assistant and build a team with a router on top.

There’s an orchestrator. Mine is named Duke. He’s the one I talk to. He doesn’t do the work himself. He routes it. When I bring him a sales question, he hands it to Buck. When I need brand strategy, he hands it to Doc. When something is messy and needs to be filed, he hands it to Betty. Each agent has its own role, its own job, and they share context.

The whole team lives as a folder structure on my computer. Plain text files. Instructions. Context. Each specialist is a few files. The orchestrator is a few more. The work moves between them through commands, not clicks.

This is the brown belt. And it’s where AI stops being a tool you use and starts being an operating layer for how the business actually runs.

It’s also where the foundation conversation from the manifesto gets real. Because at Level 4, every agent is loading context that you defined. The customer psychology. The data. The story. The brand voice. The measurement logic. If those layers are coherent, the team amplifies coherence. If they’re incoherent, the team amplifies incoherence faster than any single chat ever could.

The architecture is not the point. What the architecture protects is the point.

Level 5. The Factory

Level 5 is when the team runs without you.

You set the work. You go to sleep. Fifteen agents are running tasks while you’re not at the keyboard. You wake up to finished output. You review it. You ship the good. You send the bad back.

I’m not at black belt yet. I don’t really have much of a need to be at this point.

A few things to know before you chase it.

Tokens cost money. A team of agents running overnight on a complex project can burn through hundreds of dollars while you sleep. Sometimes that’s worth it. A team of humans would have cost ten times more. Sometimes it isn’t. You’re optimizing the wrong thing.

And there’s a deeper risk. The human-in-the-loop is the thing that keeps the work pointed at something true. Remove the human entirely and you might be amplifying incoherence at maximum speed with nobody watching. That’s not an AI problem. That’s a system problem. And the system problem doesn’t go away just because the agents got faster.

Most people don’t need to be here. Most people need to be solid at Level 4 and selective about which workflows are worth fully automating.

Where to start

Look at where you actually are. Not where you wish you were.

If you’re still in one chat that you’re afraid to close, you’re at Level 1. That’s most people. The next move is to learn projects and build one. One. Not ten. Pick the work you do most often and build the assistant for it.

If you’ve got projects working, you already know the unlock. Get into a desktop tool that can read and write your files and watch how much friction disappears.

If you’re at Level 3 and starting to feel the ceiling, that’s the signal to read about orchestration and start sketching what your team would look like.

Coherence before creativity. Diagnosis before deployment. The manifesto’s logic applies to the ladder itself.

Now you know where you are. The next move is yours.

Tell me where you are on this. I’m curious what level most people land on when they actually look.

Quick heads up before the story: I'm hosting a live Watering Hole session tomorrow on this topic. Free, open. Link at the bottom. Now to what happened on the show this week.

This week’s episode with was supposed to be about behavioral science. It ended up being about AI.

Will runs MindState Group. He built a platform called Bevy that uses behavioral psychology to help enterprise marketing teams write sharper messaging. His clients pay six figures a year for it. They are exactly the kind of teams you would expect to be running ahead of everyone on AI.

Then he said something I have not stopped thinking about.

One of his clients almost forgot to renew. The platform had nearly saved his week earlier in the year by stepping in like a chief operating officer. And he had not logged in for six weeks.

Will’s diagnosis was clean. The client was not getting less value. He was overwhelmed. He had nine AI tools open in nine tabs and could not remember which one did what. So he defaulted to none of them.

That is the pattern I see everywhere right now.

People are not drowning in AI because the tools are bad. They are drowning because they are stuck at Level 1. They never crossed the threshold where AI stops being a chat window and starts being a system.

I spent most of last year stuck there too. I treated it like Google. Open a tab, type a question, copy the answer, close the tab. Thirty good answers, zero compounding. Every conversation started from scratch because every conversation was a stranger.

Then about a year ago I built my first Project in ChatGPT. Loaded the context once. Wrote the instructions once. Came back the next day and the thing remembered who I was, what I worked on, and how I think.

Not a better prompt. Not a better model. The system itself.

That was Level 2. And once I felt the difference, I could not unsee it.

There are five levels.

Intern. Assistant. Coworker. Team. Factory. White belt to black belt, if you prefer the dojo version. Each one is a different relationship with the tool, and each one unlocks something the previous level could not touch. You climb when you are ready, not before. But you cannot climb if no one ever shows you the staircase exists.

That is why I am running a live session on this Friday.

The Bullhorns and Bullseyes Watering Hole is free. It is open. I am going to walk through the five levels, name the move that takes you from one to the next, and show what a Level 2 setup actually looks like on a screen. No theory disconnected from practice. Just the work.

I am doing this because of conversations like the one with Will. Smart, well-resourced people are losing weeks to tool fatigue. Other smart, well-resourced people are doing five times the output with calmer minds. The difference is not talent. It is not budget. It is whether someone showed them the staircase.

If you are still typing one question at a time into a chat tab, this session is for you.

If you have a Project or two running but you can feel that something bigger is sitting just out of reach, this session is for you.

If you are already running agents and orchestrating teams, you probably do not need it. Send it to someone who does.

The Watering Hole — Weekly Live Q&A Friday May 15, 1:00 PM EST
Register here:

REGISTER

It is live this week. I will share the recording with subscribers afterward if you cannot make it, but the live conversation is where the questions get answered.

See you Friday.

Curtis

P.S. The Will Leach episode is live now on the B&B feed if you want the full conversation. The AI thread runs through the back half. Worth the listen.

Tom and Curtis welcome back Will Leach, best-selling author of Marketing to Mind States and founder of the Mindstate Group, for a deeper dive into how behavioral psychology and neuroscience should shape the way marketers think, message, and measure. Building on their previous conversation, this episode moves from theory toward practice, examining why even the most intellectually convinced marketers default to feature-first thinking the moment they sit down to write copy or plan a campaign…completely misaligned with the psychology of the buyer.

This episode builds and expands upon the themes from episode 1: why marketing drifts, why mirror marketing fails, and why AI can be a false prophet if it’s amplifying the wrong things. How do you know what the right things actually are? Tune in to find out, as Will demonstrates how customers make purchase decisions…and teases our next lesson: How you can find out how your customer truly feels!

N.B.:

• Learn more and get the book

• Connect with Will on LinkedIn

• Let’s continue the conversation on our Substack! And please remember to subscribe so you never miss an episode!

Takeaways:

• The brain’s filter—the reticular activating system—blocks most marketing before it ever reaches conscious awareness; only pain points and genuine aspirations get through.

• Marketers understand consumer psychology as individuals but abandon it when thinking as providers, defaulting to feature comparisons instead.

• Short-term revenue accountability is the structural reason most marketing stays shallow; optimizing for measurable metrics crowds out deeper customer understanding.

• Large language models are trained on the world’s data, not your customer’s—what they return sounds good but is essentially a very confident average.

• Temporal landmarks—predictable moments in time when a specific mind state is likely active—let marketers target context without needing to identify individual psychological states.

• The CFO conversation changes when you can articulate what makes customers tick and predict messaging outcomes, rather than reporting last month’s clicks.

• Before any marketing meeting moves to tactics or metrics, it should start with one question: what is the customer’s pain or aspiration right now?

• Review data, call transcripts, and social media all contain psychological “tells”; AI can help surface them if it’s been grounded in behavioral science frameworks first.

• Brand is increasingly the only durable moat—not pricing, distribution, or operations—and LLMs will recognize and amplify brands that have earned genuine psychological meaning.

• AI amplifies whatever you activate—right or wrong. Getting the customer psychology right before scaling is no longer optional.

Find and Follow:

• Find all episodes at bullhornsbullseyes.com.

• Be sure to subscribe to our Substack to never miss an episode!

• Follow the show on LinkedIn!

• Learn more about Collideascope and Creative Mill at their respective websites.

• Connect with Curtis and Tom on LinkedIn.

• Check out our newsletter, Amplify and Aim!

This episode introduces a new framework for getting back on course, starting with a simple but critical shift: diagnosis before deployment.

• And please remember to subscribe the Substack newsletter so you never miss an episode!

Takeaways:

• Warning: AI amplifies whatever you feed it—right or wrong.

• Marketing drift happens gradually, then suddenly—and most teams don’t recognize it.

• AI often accelerates that drift by scaling misaligned messaging and tactics.

• The chaos tax is the cost of drift: increased effort, spend, and complexity without results.

• Most brands rely on internal assumptions instead of real customer insights.

• Diagnosis before deployment helps realign marketing before scaling it.

• The path forward starts by reconnecting with the customer—not doubling down on tactics.

• The four principles for modern marketing that we will explore in Season 3:

→ Diagnosis before deployment

→ Coherence before creativity

→ Meaning before media

→ Revenue before reach

“Two ways,” Mike said. “Gradually. And then suddenly.”

That’s Hemingway. The Sun Also Rises. And it’s the best description I’ve found for what’s happening to marketing budgets right now.

We’ve been quiet for four months. Not because Curtis and I ran out of opinions. Because we decided that if Season 3 were going to exist, it had to be something you could actually use. Not a collection of loosely connected conversations. A sequence. Eight episodes. Each one building on the last. Starting with this one. »

By the end, you’ll have something that works whether you listen once or come back to it in a year.

Think of Season 3 as an 16-week semester of higher education, with each episode serving both as a discrete learning module and as a sequential building block leading up to the final exam. There is a purposeful order to things this year. And we’re relying once again on some of the best and brightest minds in marketing and business development to help us advance the learning.

It starts with the uncomfortable part.

The wrong station, turned up loud. I used this analogy with Curtis during the episode and I want to share it here because I think it lands.

Imagine your marketing is a radio. And the radio is stuck on country music. But your customer hates country. AI comes along and says: just turn it up. Louder. More channels. More volume. And all your customer is thinking is, this is not for me. Their inclination isn’t to listen harder. It’s to change the station.

That’s the trap. Companies are spending more, producing more, deploying more. AI lets them do it all faster. And nobody is stopping to ask: are we even on the right station for the right listener?

Curtis put it bluntly in the episode. He said it’s like calling IT because your network is down, and instead of checking whether the cable is plugged in, they replace the $50,000 server.

He spent 15 years in IT before marketing, and the thing that drives him crazy is how quickly companies skip the simple diagnosis and jump straight to expensive solutions. You wouldn’t do that with a server. But companies do it with their marketing budget every single day.

Mirror, Mirror…

Here’s the part from the episode that I keep coming back to.

Mirror, mirror on the wall, who’s the fairest of them all?

You know the line. The queen asks the mirror. The mirror tells her exactly what she wants to hear. She’s pacified. Everything is fine. Until it isn’t. And when the truth finally shows up, she goes into a rage.

That’s what most marketing looks like.

Your homepage says “We are...” Your pitch deck leads with company history. Your AI tools polish that mirror until it gleams. And the mirror keeps telling you what you want to hear. You’re the fairest of them all. Your product is the best. Your differentiators are clear.

But the customer is not in the mirror. The customer is on the other side of the window. And most companies haven’t actually looked through that window in a very long time.

Curtis made a point that stopped me during our conversation. He said when you train AI without your customer’s reality, you’re just training it with the mirror. All it knows is your reflection. And a better reflection of yourself is not a reflection of your customer. Of course it’s going to tell you that you’re the fairest of them all.

But does the customer care about that? Not yet…

Nobody does this on purpose. It’s not vanity. It’s just the natural gravity of running a business. You know yourself. You talk about yourself. And gradually, then suddenly, you realize nobody’s listening.

The Chaos Tax

This is Curtis’s concept, and I’ll let him explain the math, because that’s what he does.

You know what you pay for a lead. You might know your close rate. But do you know what you pay for a lead that actually becomes revenue? Not a form fill. A customer.

The gap between those two numbers is what Curtis calls the Chaos Tax. It never shows up on a P&L. But it compounds. Every quarter you operate inside the trap, you’re not just wasting that quarter’s budget—you’re training your algorithms, your team, and your market to expect the wrong thing from you.

AI makes the slope steeper. Not because AI is bad. Because AI is indifferent. It amplifies whatever you give it. If the inputs are wrong, you just waste money faster. With better production value.

Going the wrong way. And making incredible time.

Four Gates

We laid down four principles in this episode. I’d call them sequence gates. You don’t get to skip ahead.

Coherence before creativity. This one’s mine, so I’ll own it. I love the creative side of this business. But I can’t write a tagline that works if I don’t coherently understand what I’m being creative about. A beautiful message built on the wrong insight is just a beautiful mistake.

Diagnosis before deployment. Curtis’s. He’s never going to let you spend money on ads before confirming your tracking works. Think of it like the IT help desk asking if you’ve rebooted. It’s annoying. It’s also the right first question. And, increasingly, he’s not going to let you deploy your campaigns until you’ve done the coherence work first.

Meaning before media. Talk to five customers. Find the truth. Find what actually matters to them. Then go spend money on commercials and marketing. Not the other way around.

Revenue before reach. Prove it converts before you scale it. The algorithms will find more people like the ones already buying. But only if you’ve done the work to know who those people are.

Your Homework (that I know you won’t do, but should!)

I’m giving you homework. I’m also betting exactly zero of you will do it. But here it is.

Go have actual conversations with five customers. Ask them four questions:

→ What was life like before you even heard of us?

→ What pain made you go looking?

→ Why did you choose us? You could have chosen anyone.

→ How is your life different now that you’ve been working with us?

If you do this, you’ll have the foundation for everything this season is going to teach you.

If you skip it because you “already know the answers,” well. That’s the mirror talking.

Two reasons people won’t do this:

One, they’re afraid a customer might say something they don’t want to hear.

Two, they’re convinced they already know what customers will say.

Both of those are the trap. Both of them cost you money every quarter you avoid them.

What’s on the Syllabus This Semester

We didn’t offer a single solution in this episode. On purpose. This is the diagnosis. The indictment.

Over the next eight episodes, Curtis and I are building the solution. The science behind why customers actually choose. How to excavate truth your customers won’t volunteer. How to turn that truth into a story that moves revenue. How to build a community that compounds. How to own your audience instead of renting it on platforms that can change the rules tomorrow.

By the end, you’ll have a system. Not theory. Something you can run.

But it starts here. With one honest question: When was the last time you actually talked to a customer? Not surveyed. Talked. And listened.

If the answer is more than 90 days ago, you’re marketing from memory. Not reality.

Watch or Listen to Season 3, Episode 1: THE TRAP

Available on YouTube, Apple Podcasts, Spotify, and everywhere you listen.

If this resonates, read the piece that started this whole arc: AI Isn’t Your Prophet. It Might Be Your Amplifier.

Bullhorns & Bullseyes is hosted by Curtis Hays and Tom Nixon. Season 3 drops a new episode every two weeks.

Your competitor just announced they’re “AI-powered.”

New tools. New workflows. New vendor. LinkedIn post about it with a rocket emoji.

And here’s the thing. In about nine months, their numbers are going to look exactly the same. Maybe worse. Because the part they invested in — the AI — is about 10% of what actually makes a system work.

The other 90% is boring. It’s infrastructure. Logic. Rules. Plumbing that nobody posts about because it doesn’t sound like the future.

But it is.

I’ve been thinking about this since I wrote about AI being an amplifier, not a prophet. That piece was about what happens when you turn AI loose on a broken system. It amplifies the break.

This piece is about what’s underneath the amplifier. What the system actually needs to look like before AI earns its seat.

The Pattern That Keeps Showing Up

I’ve been watching something across our clients and across every implementation I study that actually produces results. The same ratio keeps appearing.

Then I came across Jake Van Clief’s work on what he calls Computational Orchestration, and I realized he’d put a name to something we’d been doing for years without articulating it.

Jake’s research shows a 60/30/10 pattern in every successful AI implementation he’s built or observed. And when I mapped it against how we actually build client systems, it was nearly exact.

About 60% of what works is traditional infrastructure. Tracking. CRM integration. Conversion definitions. Offline conversion data flowing back to the ad platforms. Pixels firing correctly. The forms actually routing to the right person.

None of that is AI. It’s plumbing. And when it’s broken, nothing downstream works. Not your campaigns. Not your lead scoring. Not your AI-powered anything.

Another 30% is structured logic. Rules. Sequences. Decision trees. If a lead comes from this source, route it here. If the close rate drops below this threshold, check the handoff before touching the ad spend. If your sales team is rewriting the pitch before every meeting, the problem isn’t the rep — it’s that marketing made a promise the sales conversation can’t keep.

That’s not AI either. That’s architecture. Human judgment encoded into repeatable process.

The remaining 10% — maybe — is where AI actually belongs. What Jake calls the semantic layer. The part that requires understanding meaning, not just executing rules.

Reading a customer call transcript and finding the fear underneath the words. Identifying the identity tension that’s actually driving a purchase decision. Generating content that carries a specific voice because the constraints that define that voice already exist.

The 10% makes the system feel like it understands you.

The 90% makes the system actually work.

We didn’t learn this from Jake. We’d been building systems this way since before we had language for it. But his framework gave the pattern a name. And once something has a name, you can teach it. You can audit against it. You can stop having the same argument about whether AI is the strategy or the tool.

Why This Ratio Makes People Uncomfortable

It’s not a compelling pitch. Nobody raises a round on “we built a really solid CRM integration.” Nobody gets promoted for saying “I spent three months making sure our conversion definitions match between marketing and sales.”

But that’s where the money is.

Every company I’ve worked with that has a rising cost per acquisition and can’t explain why — the answer is in the 90%. Not the 10%.

The tracking is incomplete, so the ad platform is optimizing toward the wrong signal. The CRM isn’t feeding disposition data back, so marketing has no idea which leads actually became revenue. The sales qualification definitions were written two years ago and nobody updated them when the product changed.

AI doesn’t fix any of that. AI makes all of it faster.

If the signal going in is wrong, AI optimizes toward wrong faster. With better grammar. In more channels. At scale.

That’s not a technology failure. That’s an architecture failure wearing a technology costume.

Where AI Actually Earns Its Place

I’m not against AI. I use it every day. But I’ve learned — slower than I’d like to admit — that AI earns its seat only in the gaps where meaning matters and rules can’t reach.

When you need to hear what a customer is actually afraid of, not just what they said on the form. AI can process volume and surface patterns a human would miss. But only if the infrastructure capturing those conversations exists first.

When you need content that sounds like a specific human being wrote it. AI can generate at that level — but only if the voice definition, the guardrails, the narrative rules are already built. Without those constraints, AI generates from the average. And the average is where every brand sounds interchangeable.

When you need to spot the moment a system starts drifting. AI can monitor for narrative inconsistency across channels, flag when the promise on the ad no longer matches the promise on the landing page. But only if someone defined what consistency means first.

Pattern recognition. Semantic interpretation. Scale without drift.

That’s the 10%. And it’s genuinely powerful — inside a system that’s already doing the other 90% right.

The Question This Raises

If 90% of what makes marketing work has nothing to do with AI, then what are you actually buying when you hire an “AI-powered” agency?

If the answer is “we use AI to write your ads and score your leads” — you’re buying the 10% without the 90% that makes it matter.

If the answer is “we build the infrastructure, install the logic, close the feedback loops, and then apply AI only where semantic understanding is required” — that’s a system. That’s what compounds.

The companies getting this right don’t lead with the AI. They lead with the architecture. The AI is the last thing they add. Not because it’s unimportant. Because it depends on everything beneath it.

The Real Cost of Skipping the 90%

Every quarter you run AI on top of broken infrastructure, the Chaos Tax compounds. Your cost per acquisition rises and nobody can explain why because the dashboards look fine. Your sales team drifts further from your marketing because the handoff was never defined in the system. Your content gets produced faster but says less because there were no constraints to hold the voice in place.

It’s not a cliff. It’s a slope. And AI makes the slope steeper because it accelerates whatever direction you’re already headed.

Coherence or incoherence. Compounding or decay. Same tool. Different foundation.

The 60/30/10 Test

Here’s a diagnostic you can run on your own system right now.

The 60% check — Infrastructure:
Can you trace a lead from first click to closed revenue in your CRM? If someone fills out a form or calls the office today, does your ad platform know — within a reasonable window — whether that lead was qualified? Is your tracking firing on every conversion point, or are there gaps you’ve been meaning to fix?

If the answer to any of those is no, you don’t have a marketing problem. You have a measurement problem. And no amount of AI fixes a measurement problem.

The 30% check — Logic:
Do your sales and marketing teams agree on what a qualified lead looks like? Is there a documented process for what happens when a lead goes dark? When performance drops, does your team have a structured decision tree — is it targeting, offer, story, or handoff — or does everyone just start adjusting budget?

If those rules don’t exist, you’re running on instinct. Instinct works until it doesn’t. And when it stops working, nobody can diagnose why because there’s nothing written down to audit.

The 10% check — AI:
Is AI being applied to problems that require semantic understanding? Or is it being used to automate things that a spreadsheet formula or a simple integration could handle?

If your AI is doing work that doesn’t require understanding meaning — it’s expensive automation. Not intelligence.

The foundation is still the whole conversation

Last time, I wrote that the question was never “should we use AI?” It was “do we have a system worth amplifying?”

This is the next layer of that argument. The system worth amplifying is mostly invisible. It’s infrastructure and logic and rules and definitions that nobody writes LinkedIn posts about.

AI sits on top. Thin. Powerful. But dependent.

The companies that will win the next five years aren’t the ones with the best AI. They’re the ones with the best 90%.

Diagnosis before deployment.
Meaning before media.
Coherince before creativity.
Revenue before reach.

The plumbing is the strategy.

You chose the more expensive option and you can’t fully explain why.

You post-rationalized it with features you haven’t even used yet. The decision just felt right. Good, even. And yet as a provider, you’re still writing feature comparison charts and making bold performance claims. Your customer is acting exactly like you did—and you’re ignoring it.

That’s the disconnect Will Leach came back to Bullhorns and Bullseyes to explain.

The Subconscious Decides What Is Bought…and When

Will wrote the book—literally—on marketing to mind states. His research shows that the vast majority of human decisions are made at the subconscious level. If your marketing isn’t speaking to the subconscious, it will always be less effective.

Not sometimes. Always.

The reason most marketers resist this isn’t because they don’t believe it. Most of them, when pressed, will nod and say, yes, of course, emotion plays a role.

The problem is what gets measured. And what gets measured gets done.

About a decade ago, marketing became accountable for revenue delivery. If you can see it in a dashboard—cost per lead, click-through rate, conversions—it gets funded.

What can’t be easily measured is the deep psychological understanding of your customer. So that’s what gets cut.

As Will pointed out, even he didn’t get it fully right on his own book cover by his own admission. The subtitle reads: A Practical Guide to Applying Behavioral Design to Research and Marketing. That’s a feature. Nothing emotional in it. The expert wrote a feature list about the importance of not writing feature lists.

So we all do it. It’s natural to do it. Comfortable, even. But here’s how to undo it going forward.

The Tells Are There—If You Know What to Look For

One of the things Will pushed back on in the episode is the assumption that psychological mind states are invisible. They’re expressed, he said. We just don’t know how to read them.

Think of it like poker. When customers talk—in reviews, in call transcripts, on social media—there are tells. Words they choose. The way they frame a problem. What they lead with versus what they bury. AI can actually help surface these patterns at scale. Will is doing exactly that: running client call transcripts through AI trained on his behavioral framework to identify which mind states are showing up in the data.

Curtis brought a real example: a client with 16,600 phone calls from last year. They ran those transcripts through AI and found common themes that are now shaping the positioning strategy for this year. That’s not replacing the human interview. That’s using the data you already have.

But Will was clear: AI will never ever catch the quiver in someone’s voice. He’s been in the room when someone broke down crying in a closet during a customer interview. That’s the best insight he’s ever gotten. No transcript will capture it.

The Reticular Activating System

There’s a mechanism in the brain called the RAS: the Reticular Activating System. It sits at the base of the brain and acts as a filter, constantly sorting billions of pieces of incoming stimuli based on what’s relevant to your current goals and fears. If your marketing doesn’t speak to those goals or fears, it doesn’t make it through the filter. It doesn’t matter how well-produced the ad is. It won’t register.

This is why understanding the “bleeding neck problem” matters. Not the product spec. Not the differentiator. The specific pain that, if left unaddressed, is going to cause real damage. When you speak to that, you get through the filter. When you don’t, you’re invisible no matter how loud you turn up the volume.

Will’s homework for listeners: before any marketing meeting, start by asking what is the customer’s pain or aspiration right now? If you can’t answer that, stop the meeting.

Tom added his own: go to your homepage and look at the copy above the fold. If the subject of the first sentence is your company—we, our, the business name—you’re in Mirror Marketing territory.

Brand is Your Moat.

Before the episode wrapped, Will shared a conversation he’d had with a former CMO of 7-Eleven. The debate in the room was whether AI agents would make brands irrelevant—agents talking to agents, humans removed from purchase decisions.

Will’s take was the opposite. Brand will be the moat. Not operations, not distribution, not pricing. The thing that LLMs will recognize and trust. The one asset that survives commoditization.

Curtis tied it back to the season’s thesis: if you know your brand context deeply and put it into AI, AI works for you. If you don’t, it amplifies your drift.

Gradually, then suddenly.

What If You Had a Crystal Ball?

It comes down to this: What if you knew exactly WHY your next prospect would be highly motivated to buy from you? And when? And why they wouldn’t?

And what if that prospect told you exactly what they needed to hear in order to be convinced that you were the product of service of choice?

What if you could see into their subconscious and read both their minds and their movitations—to know the emotional impulse at the core of their decision to make a purchase?

What would you do with that information?

Next week, Emily Bielak is going to return to the podcast to show you how to get it.

👉 Watch or Listen to Season 3, Episode 2 here.

Available on YouTube, Apple Podcasts, Spotify, and everywhere you listen.

Bullhorns & Bullseyes is hosted by Curtis Hays and Tom Nixon. Season 3 drops a new episode every two weeks.

I need to say something that might be uncomfortable.

You can generate a full marketing campaign in about 45 minutes now. Copy. Images. Landing pages. Email sequences. Ad variations. All of it.

And here’s the part nobody wants to say out loud. It will look professional. It will sound competent.

And it will not work.

Not because the AI failed. Because the system underneath it was never aligned in the first place. You’re just running that misalignment at 100x speed now. With better production value.

The appeal is real. I get it.

Look, if you’re a founder, a CEO, a marketing leader, the promise of AI is almost impossible to ignore. Not enough time. Not enough budget. Not enough people. And here comes this thing that says: I’ll write your ads. I’ll score your leads. I’ll personalize your emails. I’ll analyze thousands of customer comments in seconds.

It feels like the answer. I understand why people grab onto it.

But there’s a cost hiding behind that promise. And it doesn’t just add up. It compounds. Every quarter you lean on this without a foundation, the problem gets deeper and the price gets steeper.

Here’s the math most people skip.

AI is an extraordinary amplifier. But if your fundamental marketing system is broken, if you don’t actually know your customer’s internal reality, if your messaging says one thing on the website and something different in the sales deck, if your sales team is rewriting the pitch before every meeting because the marketing doesn’t match what they’re hearing in conversations. All AI does is help you do that wrong stuff faster.

You burn through money faster. But the emails look great.

The mirror problem

Most companies, probably without realizing it, are doing what we call Mirror Marketing. Their homepage says “We are...” Their ads say “Our product does...” Their decks lead with company history and awards. Their AI tools just polish that mirror until it gleams.

But the mirror is still pointed inward.

What actually moves cost curves without increasing spend is the opposite. Looking out at the customer’s world. Speaking to their reality. Their tension. Their language. Not yours.

AI doesn’t know the difference. It generates from the statistical average of everything in your category. And the average is exactly where revenue goes to die.

The Chaos Tax

Here’s something that changed how I think about this.

You probably know what you pay for a lead. You might know your close rate. But almost nobody I talk to knows what they pay for a lead that actually turns into revenue. Not a form fill. Not a demo request. A lead that becomes a customer.

The gap between those two numbers is what we call the Chaos Tax. It’s the most expensive line item that never shows up on your P&L.

And AI makes it worse. Not because AI is bad. Because AI is indifferent. It amplifies whatever system you give it. If the inputs are misaligned, wrong customer psychology, wrong narrative, wrong measurement infrastructure, AI just amplifies the misalignment. At scale. With perfect grammar.

So what does AI actually need?

Not a better prompt. A foundation. And the order matters because each layer depends on the one before it.

Psychology first. Understanding why people actually choose. Not features. Not price comparisons. Identity protection. Fear. Aspiration. The stuff that never shows up in your click-through data. AI gives you the average. Revenue lives in the specific.

I used to skip this step. I came from the data side. I wanted to go straight to what I could measure. It took me longer than I’d like to admit to understand that the measurable stuff only works when the psychology underneath it is right.

Then data. Not spreadsheet data. Excavated customer truth. The pause before someone names their real fear on a call. The tension between what a customer says they want and what their behavior reveals they need. AI can process volume and surface word frequency. It can’t hear what’s underneath the words.

Then story. And here’s the trap. You can generate an entire messaging framework in 20 minutes now. Mission, vision, value propositions, taglines. It’ll be grammatically flawless. Tonally appropriate. And completely interchangeable with your three closest competitors. Because AI generates from the average of all messaging in your category. If your brand story could have been written for anyone in your space, it’s not a story. It’s furniture.

Then community. AI can simulate warmth. It cannot generate belonging. Belonging requires risk. Vulnerability. The possibility of rejection. The most human company wins. And AI is not human. I resisted this idea for a long time. I’m a systems guy. But the data kept telling me the same thing. People buy from people they feel connected to. No algorithm changes that.

Then owned audience. AI makes it trivially easy to produce more content for social platforms. More posts. More videos. More carousels. Faster, cheaper, higher volume. And every piece of that content builds on someone else’s land. You’re producing more stuff, faster, on platforms you don’t control. Volume isn’t the problem. Ownership is the problem.

Then measurement. Closed-loop. Revenue-connected. Not dashboards that tell one story while the actual revenue tells a different one. AI is being used for lead scoring, automated follow-up, chatbot qualification. All useful if the definitions underneath are correct. But if your MQL definition is wrong, AI just scores wrong leads faster.

That’s the system. Psychology. Data. Story. Community. Audience. Measurement. In that order. Each layer protecting the one before it.

Remove any one and the others weaken.

The resolution

AI as a false prophet, this belief that tools can replace understanding, is the most expensive seduction in marketing right now.

But AI as an amplifier, applied to a system that actually works? That’s where it gets genuinely powerful.

When the foundation exists, AI can:

• Enforce your brand voice at scale. Because you’ve defined what your voice actually is. Through real excavation and real narrative architecture. Not a brainstorm on a whiteboard.

• Generate content that sounds like you. Because constraints exist to prevent drift.

• Accelerate research synthesis. Because you know what questions to ask before AI summarizes the answers.

• Optimize campaigns faster. Because your measurement infrastructure captures the right data and feeds it back.

• Personalize at scale. Because you’ve defined psychological clusters, not just demographics.

What AI cannot do:

• Feel your customer’s identity tension.

• Know what your customer is afraid to lose.

• Build belonging.

• Create the kind of human vulnerability that earns trust.

• Replace lived experience with statistical probability.

• Generate gravity from nothing.

AI is agnostic. The system determines the outcome.

If you amplify coherence, you get compounding growth. If you amplify incoherence, you get compounding waste. Same tool. Different results. The variable is what you built before you turned it on.

The question was never “should we use AI?”

The question was always: do we have a system worth amplifying?

Every quarter you operate without that system, the Chaos Tax compounds. Your CAC rises. Your messaging fragments. Your sales team drifts further from your marketing. Your AI tools amplify the incoherence and nobody can figure out why the numbers aren’t working because the dashboards all look fine.

It’s not a cliff. It’s a slope. And the slope gets steeper.

The tool is not the problem. The tool is not the solution.

The foundation is the whole conversation.

Coherence before creativity. Diagnosis before deployment. Meaning before media. Revenue before reach.

That’s the operating logic.

And no algorithm changes it.

Two years ago, simply showing up was enough.

If you published consistently, demonstrated expertise, and shared thoughtful perspectives over time, you separated yourself from most of the market. Content marketing worked because not everyone was doing it well — or doing it at all.

Today, everyone is showing up.

And most of them are using AI to do it.

That doesn’t mean content marketing is dead. It means generic presence is dead. The problem isn’t volume. It’s believe-ability.

We are no longer just competing with other professionals. We are competing with artificial voices, auto-generated expertise, and an endless stream of frictionless content. Even if your ideas are legitimate and human-generated, your audience may assume they aren’t. And if they believe they can prompt the same answer from ChatGPT in ten seconds, why should they read your blog post?

That is the real shift.

The End of Serial Publishing

For years, the standard prescription looked like this: pick a topic, write a post, publish it, and repeat next week. Over time, that steady drumbeat built authority and brand preference. When someone eventually entered the market for your services, you were top of mind.

In theory, that still works. In practice, it now produces something far less satisfying: invisibility.

When every piece of content stands alone, when each post resets the clock, and when nothing builds toward a defined outcome, you are not creating momentum. You are maintaining motion. The difference matters.

If 2026 demands anything of content marketers, it’s this: stop publishing serially and start building strategically.

Instead of thinking in weekly posts, think in 90-day arcs. Instead of asking, “What should we write about this week?” ask, “What outcome are we driving toward this quarter?”

Content without a defined destination is now just noise with grammar.

The Four “Purrs”

If your content strategy is going to cut through AI saturation and audience skepticism, it needs four deliberate characteristics. Together, they create what I call The Four “Purrs” — and ultimately, the fifth and most important result. (Spoiler alert: it also rhymes with purr.)

1. Purposeful

Content must be aimed. Not vaguely. Not aspirationally. Specifically.

It’s no longer enough to say, “Over time, this will build trust.” The strategy should define a measurable objective inside a defined window — typically 90 days.

For example: one clear purpose is to enroll as many members of your addressable market as possible into a live webinar, workshop, or in-person event. That singular objective changes how you create content. Each post becomes runway. Each video becomes reinforcement. Each email becomes part of a sequence.

You begin to see your audience more clearly as well. There are learners who consume quietly. There are lurkers you cannot measure but know are there. And there are hand-raisers — the persuadable segment who are willing to take the next step.

Your content must give them something to raise their hand to. Purpose creates direction. Direction creates momentum.

2. Persistent

Consistency used to be enough. Now it must be consistency with intention. Persistence is different: it’s consistency with purpose.

Persistence means narrowing your focus over that 90-day arc. Instead of writing about a dozen disconnected topics, you select one core tension: one problem, one theme, one drumbeat …and you’re going to approach it persistently from multiple angles.

You explore it through stories, analysis, examples, frameworks, and commentary. You repeat the central idea without repeating yourself. By the time your culminating event arrives, your audience should not wonder what it will cover. They should be expecting it.

Persistence builds familiarity. Familiarity builds trust. Trust makes persuasion possible.

3. Persuasive

Education is valuable, but information alone is no longer scarce. Transformation is.

Professional services are typically purchased at a moment of inflection — when someone decides they can no longer tolerate the status quo or they finally choose to pursue an aspiration. Effective content moves people closer to that decision.

This does not require manipulation or aggressive selling. It requires clarity. Your content should demonstrate that the current path has limitations, that an alternative exists, and that you understand the emotional and practical tension your audience is experiencing.

If your strategy is purposeful and persistent, persuasion becomes a natural byproduct. You are not pitching constantly; you are preparing thoughtfully. Over time, the persuadable segment of your audience begins to consider change.

4. Personal

This is where the human advantage becomes unmistakable.

Written content alone now competes directly with AI. Machines will always win on speed and scale. Humans win on presence.

Video. Live webinars. Unscripted commentary. Podcasts. Real conversations. When people see you think in real time, when they hear your cadence and observe nuance in your delivery, your expertise becomes authenticated. You signal authorship. You signal authenticity.

In an era of synthetic content, visible humanity becomes a differentiator.

Personal content does not replace written content. It validates it.

The Fifth “Purr” → Permanence

When content is purposeful, persistent, persuasive, and personal, something powerful happens: it gains permanence.

Instead of disappearing into the feed after 24 hours, your ideas stay in the market for 90 days. They build anticipation toward a defined moment — a live event, a webinar, a workshop, or a briefing.

That event then becomes fuel for the next 90 days.

You record it. You gate the replay. You extract short-form video clips. You generate companion blog posts. You develop follow-up emails. You create derivative assets that extend the lifecycle of a single idea.

Your strategy compounds rather than resets.

You are no longer producing isolated posts. You are programming reputation over time.

The New Discipline

The truth is, AI did not kill content marketing. It exposed outdated strategy.

The brands that will win in 2026 and beyond will not be the ones producing the most content. They will be the ones building structured, human, outcome-driven arcs that culminate in meaningful engagement.

The question is no longer, “What should we publish this week?”

It is, “What are we building toward over the next 90 days?”

Answer that well, and your content engine won’t sputter. It will purr.