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.

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.

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.

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.

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%.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.