What this covers
Tiago Forte interviews serial entrepreneur Hayden, who has built roughly 30 active businesses across brick-and-mortar and digital ventures, to introduce the master prompt method — a technique for loading AI with persistent, comprehensive business context so that every query benefits from full company knowledge rather than starting from scratch. The core argument is that context quality transforms AI from a generic search tool into a strategic operating system, improving output quality by roughly 40 percent and collapsing months of work into hours. Hayden walks through the rationale, demonstrates the method in practice, and offers a framework for building a first master prompt.
The video covers the mechanics of how persistent context works in different tools, arguing that only Claude currently supports the method effectively through personal preferences and shareable projects, while ChatGPT lacks project sharing and Gemini's equivalent feature does not function reliably. Hayden sketches what this shift enables: narrowing the execution gap between small and large companies (letting a one-person operation run processes like top-grading hiring that once required entire teams), flattening organizational hierarchies by eliminating director-level roles, and potentially doubling to tripling growth rates across industries — though at different paces, with digital businesses moving faster than brick-and-mortar. He also addresses where AI reaches its limits, noting that sales work is bounded by human call-taking and that strategy should be left out of early prompts because AI is shifting what's possible so rapidly that today's plans will change radically. The practical section walks through structuring a master prompt document with sections for personal strengths, company details, market information, team members and their KPIs, products, and culture.
Hayden argues that loading AI with persistent, comprehensive business context via a reusable 'master prompt' transforms it from a Google-like assistant into a strategic operating system that democratizes execution, letting small companies perform processes once reserved for large enterprises and potentially doubling or tripling annual growth.
- AI output quality and speed depend on the amount and quality of context supplied, so a standing 'master prompt' that injects full company context into every query dramatically improves results.
- This democratizes execution, narrowing the capability gap between small and large companies, and will reshape org structures toward fewer management layers.
- Tool choice matters: Claude's personal preferences and shareable projects make the master-prompt method feasible where ChatGPT and Gemini currently fall short.
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
AI will produce a 'democratization of execution' — narrowing the gap between small and large companies by letting even a one-person company perform thorough processes (like top-grading) previously reserved for very large firms — but the pace varies by industry: digital, software, marketing, legal, and finance move fast while brick-and-mortar takes several years.
“there will be a democratization of execution. So and it's going to take a different amount of time depending on what industry you're in”
Documented processes (SOPs / fulfillment engines) drive accountability because work fails for only a few reasons — people didn't know it needed doing, didn't know how to do it, or weren't bought in — and clear documentation eliminates the first two while making the third easier by helping employees understand their role and what they're helping achieve.
“if something doesn't get done, there's only a few reasons why that happens, right? um either they didn't know it needed to be done um they didn't know how to do it or they weren't bought in”
AI will eliminate the director layer of organizations entirely while expanding the manager layer, helping companies bypass the 'valley of death' growth stage; the future structure will be far larger companies with far fewer employees — essentially owner plus front-level people, with AI serving as the director and CEO layers.
“I think there will be no director layer at all. Um and I think the manager layer will probably expand”
Using AI with the master prompt to run top-grading hiring produces a higher proportion of A-players, and because hiring success compounds, two near-identical companies will diverge dramatically over years — if Hayden gets two of three hires working out versus a competitor's one, the resulting hiring base compounds into a large gap.
“let's say 50% more people in this method are a players... you maybe need to make if you're growing maybe year one you make three hires. Um maybe one of yours works out and two of mine work out.”
A simple one-sentence prompt with no context yields answers no better than a Google search, whereas giving AI as much detailed instruction and context as possible produces dramatically better, faster, and higher-quality output — quality improving around 40%.
“If you do a simple one-s sentence prompt and you've got no other context, the answer you're going to get is going to be no better than really a Google answer in most cases.”
The master prompt should contain not just factual business information but reusable 'protocols' — trigger phrases like 'AI hiring', 'AI SOP', 'AI CMO', and '131' — that act like mini programs invoking detailed decision-making and output frameworks automatically.
“They're sort of like almost like mini programs that you write protocols. Yeah. They're like little protocols.”
Goals and strategy should be left out of a first master prompt because AI is shifting what's possible so fast that today's plans will change massively — Hayden has shortened planning from three years to six months, and his 2026 revenue target is now roughly four times what it was when set, mostly because of AI.
“the only reason I wouldn't put this in in your first master prompt is as you use AI, any goals and strategy you have today are most likely going to change massively.”
When building a master prompt, you should always ask Claude to ask you questions and answer its own questions based on what it knows, then stop for your review — this saves typing, reveals what the AI misunderstands, and lets you correct the master prompt in an iterative loop.
“when you're using Claude, I always encourage you to ask Claude to ask you questions. And then if you have a master prompt, I always encourage like answer your own question based on what you know and then stop and let me look at it”
Hayden splits his businesses into two portfolios: one that won't be replaced by technology but won't scale at 5x a year (in-person fulfillment like plumbing/HVAC, limited by human hours and physical resources), and another that can scale like crazy (info products, even sold without sales) but will face competitors also scaling fast.
“we have two portfolios. One portfolio is like will not be replaced by technology but will not scale at like 5x a year and then we have the other portfolio that's like scale like crazy.”
A first-version master prompt should be structured as a Google doc with sections for personal info (role, strengths, weaknesses, how you want AI to help), company info, market information (competitors), people/team plus each person's one KPI, products and services (costs, features, benefits), and culture (core values, mission, BHAG) — with a paragraph or one-to-two sentences per item being sufficient to start.
“I would first probably just create a Google doc. Um, and I'd start creating some sections. And so I think the first section would be personal info.”
The master prompt method is to capture full business context (sometimes a 20-30 page document) in a tool's persistent settings so that context is automatically accessed on every prompt without re-typing — Hayden built this by exporting ChatGPT's accumulated memories into Claude's account preferences and iterating from there.
“What the master prompt is was basically how can I give full context for in my case for my business every single time I prompt without having to write anything.”
A global master prompt should be paired with 'projects' that act as protected silos of context (e.g., separate seller-side vs buyer-side knowledge), because including everything in the global prompt could confuse the AI and cause it to pull the wrong information.
“there's no reason for me to put this in the master prompt that all people on the buyer side of the business will use. Might even get confusing and pull the wrong info.”
AI with the master prompt collapsed work that previously took months — an AI hiring/job-description package would have taken an HR director and CEO the better part of a week with multiple meetings, and documenting fulfillment-engine SOPs that originally took months and many people now replaces at least three months of work.
“This would have taken our HR director a bunch of time to create... this would probably take the better part of a week”
Among current tools, only Claude and Notebook LM support the master prompt method effectively — Claude via persistent personal preferences plus shareable projects with project knowledge — while ChatGPT cannot share projects or support a master prompt and Gemini's 'gems' don't work properly.
“Chat GPT can't do it. Gemini has gems and they just don't work properly... the only two that are kind of options for this are are Notebook LM and Claude.”
AI can potentially double to triple every one of Hayden's companies per year; whereas leadership teams used to set sustainable growth goals of 30-50% year-over-year, scalable companies can now achieve roughly 10x that with AI, while less scalable brick-and-mortar businesses can reach 60-100% growth per year.
“I firmly believe that AI can potentially double to triple every single one of my companies per year. That's how big I think it is.”
On the actual sales side, AI can help with call reviews, pipeline management, and autoresponders, making a sales manager roughly three times more effective by removing reporting work — but it currently cannot take calls, so output is still bounded by humans taking calls, though that may change in the next year or two.
“sales today you on the actual sales side AI can help a lot with things like call reviews, pipeline management, um autoresponders. It can make probably a manager maybe three times more effective.”
You can effectively 'install' a book or a thinking style into your business by adding it to the master prompt — e.g., prompting the AI to respond as a particular business person would, or applying a concept from a book directly into operations.
“sometimes you're reading a book, you're like, I wish I could just like install this book like a software program into my business and now you can.”
Acquire helps people become acquisition entrepreneurs by buying roughly million-dollar-a-year profitable cash-flow businesses financed through SBA loans, while Empower handles the seller side via exit planning, connecting buyers and sellers.
“Acquire is a business that helps people basically become acquisition entrepreneurs, typically buying a million dollar a year cash line business that's in profit. um financing them through SBA loans.”
Hayden spends roughly the first five hours of each day (waking around 4:30) learning and tinkering with AI, and over the past decade has spent about 40 hours a week learning something, systematizing what works, and teaching it to his leadership teams.
“typically wake up around 4:30 and spend the first 5 hours or so of the day just learning and typically playing with AI.”
Hayden operates three holding companies containing roughly 30 active businesses, about two-thirds acquired and one-third started, each business and each holding company having its own CEO, spanning brick-and-mortar (HVAC, plumbing, roofing) and digital ventures.
“I have three different holding companies... we've got probably like 30 active businesses. I'd say twothirds of those were acquired. Onethird of those were started.”