You just agreed to pay for a tool, an agency build, or an AI strategy deck described as proprietary, custom, or built specifically for you. Or maybe you’re spending time and money in tokens to build it yourself. Before that invoice clears, ask what you actually bought/built.
In 1999, the record industry was selling a file. Napster made that file free to copy. The industry’s first move was legal, not commercial: the RIAA sued that December, an injunction landed in March 2001, Napster shut itself down that July, and by June 2002 it had filed for bankruptcy. Two and a half years, mostly lawyers, and the file had already gotten away.
Software is having the same year right now. In October 2026, Kotaku’s Lewis Parker reported on old games, including Halo: Combat Evolved and Grand Theft Auto: Vice City, running in browsers after being pulled apart and rebuilt from the files that ship to players. Parker called it “IP protection whack-a-mole on a scale I don’t think we’ve seen since the advent of pirated DVDs.” That’s 1999 again: a technology outran what rights holders could control, fast. The limit this time: it works on what ships to a customer’s machine. It doesn’t reach a company’s own servers.
Napster duplicated the recording. This makes it cheaper to reconstruct software from what has already been distributed. The techniques and legal questions differ; the economic threat is familiar. Either way, what you paid for stops being rare once a working copy exists outside your control.
In McKinsey’s 2026 State of AI survey of 1,719 respondents, 32% said their organizations had decided against buying at least one software product or feature because they could build it with AI coding tools instead. That’s a purchase avoided, not a subscription canceled. It isn’t 32% of software spend disappearing, and it isn’t a SaaS collapse. More than three in ten organizations are telling a researcher, in writing, that their build-or-buy answer has shifted.
I don’t have a study on what comes next (Jake will probably tell us the answer is somewhere in our history), but the same question extends beyond software: an agency’s deliverables, a strategy deck an AI wrote for you, any procedure you paid someone to document. Ask whether a rival could get hold of a version of what you bought.
A rebuilt tool might cost less to stand up than the subscription it replaces, but someone still has to secure it, maintain it, and keep it working when the underlying platform changes. Cheap to build isn’t the same as cheap to own, whether you build or buy.
Either way, the question to ask before build-or-buy isn’t which one to pick. It’s what problem you’re actually solving. Skip that step, and you won’t know afterward whether it worked.
What the customer wants hasn’t moved. Nobody hires an agency, buys a tool, or pays for an AI build because of the file underneath it. They pay because they expect results, and they stay because they get them. Once the file behind the result is cheap to copy, you stop paying a premium for it. What you’re still paying for is whatever keeps it working: who’s accountable when it breaks, who keeps testing it, and who updates it when conditions change.
Custom, proprietary, or hard to copy doesn’t make a tool valuable. It becomes valuable because it makes a difference in the business.
I’ve written before about which part of a build is actually yours and which part is easy to hand to the next vendor (see “The Tool Is Ten Percent“); the short version holds here too.
This cuts against us as much as anyone. In June, I posted our own operating map publicly on Skool: folder structure, routing rules, and how we run client work end to end. I said plainly that I was giving it away, and that the map was never the moat. People read it. Some copied pieces of it. None of them copied ME.
Written-down knowledge also lets you run a business without carrying everything in your head, training someone new, or keeping it going when someone leaves. It’s necessary. It’s also easy to copy: playbooks, prompts, procedures, anything that tells a person or an AI how to act, including ours. A rival can take any of it tomorrow, read it, run it. What they can’t take by taking the file is the test behind it: whether that playbook still holds up against this month’s results, for your actual customers, right now. A file is a snapshot. A rival who copies it still has to run it against their own numbers before they know if it works for them.
A copied playbook can now do more than get used once. It can go straight into somebody else’s AI and tell that AI what you believe. That’s a faster path for a generic process to end up running inside a competitor’s business, word for word.
Even then, holding the file isn’t the same as knowing when to stop trusting it. A market shifts, a client’s situation changes, and last quarter’s playbook stops fitting this quarter’s problem. Copying the file doesn’t copy the discipline of testing it. Keep testing, and you’re more likely to notice when a playbook stops working, whether you wrote it or copied it. Stop testing, and the file is all you have left, ours included.
Pick anything you’re currently paying for: a tool, an agency retainer, an AI-built process. Ask two questions. First, if a rival had the exact files tomorrow, what would they still be missing? Second, what actually happened because you used it, not what it promised?
Then name one number you already use to judge the business: close rate, cost to win a customer, your marketing efficiency ratio, whatever you actually check. Write down where it sits today. Set a date to check it again, and check whether it moved in the right direction, and whether what you bought had anything to do with it.
If it moved for a reason you can trace back to what you bought, the purchase earned its price. If it didn’t, or you can’t trace the difference, ask the harder question, and don’t stop at the vendor who sold you the proprietary system. It applies to us too. If you’re paying Collideascope and can’t point to what moved and why, ask us exactly what you’d ask them: what exactly did you buy?
Beautifully put. Wise words.