The problem
Ask a generic AI for a market-entry analysis and you’ll get something that reads like one — headings for “market size,” “competitive landscape,” “go-to-market plan.” It’s fluent, fast, and it will happily cite Ansoff or Porter by name if you ask it to. What it won’t do is let those names constrain anything: the framework is decoration on top of an answer the model would have produced anyway, and two people asking the same question on two different days get two different, equally confident structures.
The fair challenge, stated plainly: can’t anyone get something that looks like rigorous analysis from AI in thirty seconds now? Yes. AI made producing something that looks right free. It did nothing for proving it is right, or getting the same standard from two people on two projects. What’s scarce isn’t the words — it’s the method, applied the same way every time, with a trail that shows the applying happened.
The canon shaped the concept, not the reverse.
The six-question route is Ansoff’s growth matrix, walked rather than read. The reconcile step — size a market four ways, then triangulate to one number on the record — is Kahneman’s outside view, operationalized instead of quoted. And checks-and-commitments — judgment at the moments money actually moves, nothing between the analysis questions — is Cooper’s go/kill discipline, stripped of the paperwork it was never really about.
Two different jobs, two different sources
Not every framework is doing the same work, and keeping that distinction honest is what stops a canon section from turning into name-dropping. Architecture sources shape the whole product — Ansoff routes the move, Kahneman governs how the engine reasons under uncertainty, Cooper shapes where a commitment sits. Those three don’t answer any single question; they decide the shape every question is asked inside.
Question sources answer one stage each, and stay scoped to it. Jobs-to-be-done (Christensen, Ulwick) structures “is the need real” — the first question, not the whole journey. Porter’s five forces structures “can we win it.” Forbis & Mehta’s Economic Value to the Customer — the actual academic origin of EVC, not a McKinsey slide — structures the pricing question. McGrath & MacMillan’s discovery-driven planning structures the assumption register that runs underneath all of it: their finding was that new ventures fail because “the ratio of assumptions to knowledge is high,” and the fix was making every assumption a named, checkable line rather than an implicit one buried in a spreadsheet.
The practical test, applied while building each stage: would this framework still be true if the product looked completely different? Ansoff, Kahneman and Cooper pass — they’d shape a pen-and-paper version of this just as much as the software. The question-level sources wouldn’t survive a redesign of their one stage; that’s exactly what marks them as scoped, not foundational.
Sources
This is one piece of the same trust story behind every AI feature in this tool — you always see exactly what would be sent, and the math is never something the AI invents.
Try the Market Size Calculator →