Apple told a court this month that OpenAI stole its secrets. The filing reads like a heist. An Apple engineer left for OpenAI, and Apple says he kept getting back into its files through a login bug, at least five times, then walked out with thousands of pages. One of them described a product Apple never announced. He was already working at OpenAI the whole time. Apple counts fourteen of its former people now sitting inside OpenAI, and it wants a judge to search every OpenAI device, even the files someone already deleted. None of this is proven. The judge hears the case on October 1.
I posted the ten craziest facts from that filing last week and it did 260 reactions and 112 comments, most of them arguing about where the line sits between hiring talent and stealing it. That is the fun question. But it is not the one that stuck with me. The part I keep thinking about is simpler and more uncomfortable. A company worth trillions could not keep its most important knowledge from leaving through the front door, inside a person's head and a person's laptop.
I run Growth Cab, a go to market advisory. Founders hire us to fix how they sell contracts above fifty thousand dollars. So I spend most of my week looking at what a company's real advantage is and where it actually lives. And the Apple story is the cleanest example I have seen this year of a mistake I watch founders make the moment they start shopping for go to market consulting services. They pay for a playbook. A playbook is the one thing that walks out the door.
in
“Apple says OpenAI stole its secrets. Here are the 10 craziest facts from the new filing:”
What Go to Market Consulting Services Actually Sell
Most go to market consulting services deliver a document. A sharp one, usually. An ideal customer profile, a positioning statement, a messaging hierarchy, a cold email sequence, a discovery script, a slide that maps the whole funnel. I have written plenty of these. They are genuinely useful in the first month. The problem is what they are made of. A document is a snapshot, and a snapshot is copyable.
Think about it the way Apple's lawyers are thinking about it right now. If your entire advantage is a framework, then your advantage has legs. Your best rep can screenshot the sequence on their last day and take it to a competitor. The consultancy that built your deck will sell eighty percent of the same deck to the next founder in your category, because that is how the model works. The secret framework was never secret. It was a rental. You felt like you owned a moat, and what you actually owned was a PDF that three other companies in your space are also reading this quarter.
This is why so many expensive GTM engagements feel great for six weeks and then quietly stop mattering. The founder got a smarter version of what they already half knew, printed it, and nothing about how the company operates changed. The knowledge sat in a file instead of getting wired into the work. And a file, as Apple just learned in public, is the easiest thing in the world to lose.
The Advantage That Cannot Be Copied
The thing worth paying for is the opposite of a document. It is an operating system. It is the set of feedback loops that turn every sales conversation your team has into something the next conversation can use. Your competitor cannot buy that from a consultant, because it is built out of your calls, your objections, your win and loss reasons, your enrichment data, your positioning that got sharper because you ran it against forty real buyers and watched which line made them lean in.
I wrote a whole piece a few weeks ago arguing that your data is the one advantage AI cannot copy, and the Apple case is the same lesson from the other direction. What Apple could not protect was static. Documents, specs, files. What it could never lose, and what actually keeps it ahead, is the compounding machine of thousands of engineers shipping and learning together every day. You cannot subpoena that. You cannot screenshot it on your way out. It only exists while it is running.
So when a founder asks me what good go to market consulting services should leave behind, my answer is not a better deck. It is a system that keeps producing after we leave. The messaging should live inside the CRM fields your reps fill on every call, instead of sitting in a slide. The ICP should be a query that refreshes on its own, so it never rots into a paragraph someone wrote once. The sequence should be tied to a data source that tops itself up, so the campaign does not starve the moment the first list runs dry. The goal is to make the advantage impossible to remove, because it is stitched into how the team already works.
How We Build It In Practice
Here is the concrete version, because principles are cheap. When we start with a client, we do not open with strategy. We open with instrumentation. We wire the CRM so that every call captures the same four things in the same fields, which means that after a month the company has structured data about why deals move instead of a folder of call recordings nobody opens. We connect enrichment so the target list refills on a trigger instead of on someone remembering. We turn the positioning into a set of standing rules the reps and the AI tools both read before they write anything, so the message stays consistent even as people rotate in and out.
That last part is the whole point. People rotate in and out. Reps leave, founders get busy, the person who understood the strategy takes a new job at your biggest competitor. If the advantage lived in that person, it just left. If it lived in the system, it stayed. The Apple engineer took thousands of pages with him. He could not take the thing that makes Apple hard to beat, because that thing is not stored anywhere he could reach. That is the standard a real go to market engagement should aim for.
Where Go to Market Consulting Services Stop Working
I would rather tell you where this breaks than pretend it always holds. Go to market consulting services cannot help you if you have not found product market fit yet. There is nothing to systematize when you do not yet know who buys and why, and building loops around a guess just makes you faster at doing the wrong thing. In that stage the founder has to be in the calls, learning by hand, and no operating system substitutes for that.
It also does not work when the founder wants to buy their way out of selling entirely. The best engagements make the founder more effective inside the motion rather than absent from it. If the goal is to never touch a sales call again, a consultant becomes an expensive delay before the same problem comes back. And it does not work when a company treats the deliverable as the finish line. The document is the start. If nobody wires it into the daily work, you have bought a very nice snapshot, and you already know what happens to snapshots.
The honest takeaway from watching Apple try to claw its secrets back through a courtroom is that the secrets were never the moat. The moat was the machine. When you evaluate go to market consulting services, ask one question. When this engagement ends, does my advantage leave with the people, or does it stay running in the system? Pay for the second one. It is the only kind that compounds, and the only kind nobody can take.
I write more of these breakdowns in AI Frontier, my newsletter on where go to market and AI actually meet. And if you want the raw version, I post it first on LinkedIn, arguments in the comments included.
