Vibe coding means anybody can ship. AI means anybody can build. A factory in Shenzhen means your physical product gets cloned in weeks. Whatever advantage your product has today is a short-term advantage. The one thing nobody can copy is the way you and your founding team solve the customer's problem. The trouble is you can't explain it, because you don't know you're doing it. There's a science for that. It's called cognitive task analysis, and it was built by Gary Klein, Robert Hoffman and Richard Clark to answer a question about firefighters: how does a fireman standing in a burning house suddenly know the floor is about to collapse? He gets his crew out. Seconds later the floor goes. Then he writes the training manual and leaves that part out, because he never knew he knew it. In this episode I take that field and point it at a founder. I walk through the four tools of applied cognitive task analysis - the practitioner's version Laura Militello and Robert Hutton published in 1998 - and show you how to run each one on yourself or your co-founder. Then I add the step the cognitive scientists don't take, because they're not in marketing: you name what you found, and you become known for it. What we cover: • Claude Hopkins, the Schlitz brewery, and "everybody does that" • Lucky Strike's "It's toasted," and the limit of saying what everyone does • The bread machine that couldn't get the crust right until an engineer noticed the baker was twisting the dough • The firefighter who said "we've got to get out of here now" • Naturalistic decision-making: why experts decide by feel and can't teach it • The four ACTA tools, in plain English • The task diagram: customer's problem, on the left. Problem solved, on the right. What are the 3 to 6 steps? • The knowledge audit: what do you see that the industry misses? • "What shortcut do you take that would look wrong to an outsider?" • Interviewing the founder and the co-founder separately, and what agreement vs disagreement tells you • The realtor who prices low on purpose and drives the price higher • The extra column: what does the industry call this, and what should it be called? • Every unnamed row is a coinage candidate • Becoming "the something guy," and where the term has to appear 📬 Newsletter: https://newsletter.viralgeniusinstitute.com 🌐 https://labastida.com/about CHAPTERS 0:00 — Differentiating a company is harder than it has ever been 1:00 — Product advantages are short-term now 1:20 — Naval Ravikant: you are the product 1:40 — Claude Hopkins and the Schlitz brewery 2:20 — Mad Men, Lucky Strike, and "Toasted" 3:20 — 1917: the real version of the story 5:00 — The second way: tacit knowledge 5:20 — Matsushita, the bread machine, and the baker's twist 6:20 — Cognitive task analysis, and naturalistic decision-making 7:00 — The firefighter who knew the floor was about to collapse 8:00 — Deciding by feel vs deciding by weighing options 9:00 — Why CTA exists: teaching what the expert never noticed doing 10:00 — The methods, and the problem with all of them 11:00 — Militello and Hutton, 1998: applied cognitive task analysis 11:20 — Tool 1, the task diagram 11:40 — Tool 2, the knowledge audit 12:00 — Tool 3, the simulation interview 12:30 — Tool 4, the cognitive demands table 13:00 — What this means for marketing 13:40 — Paul Graham, and doing things that don't scale 14:00 — Why every AI startup has a different feel for the same problem 15:00 — Running it on a founder: the task is how you diagnose the customer 15:40 — The knowledge audit questions 16:00 — The realtor and the bidding war 17:00 — Pricing low on purpose 17:40 — Founder and co-founder, separately 18:00 — The extra column: what should this be called? 19:00 — Build the framework. Draw it. Name it. 19:40 — Joe Davis, the cloud science guy 20:20 — Where the term goes: About page, LinkedIn, podcasts, stages
