Track map
Detailed content
🩹 Need and shelf life
Before you ask "how do I build it?", ask "does this need to exist?" and "will it still need to exist a year from now?" These are the two questions that eliminate the most ideas, and that's a good thing.
Module topics
What it is
The first opportunity question: does the product come from real demand, or just the excitement of being able to build it?
Why learn it
Today, you can build almost anything in an afternoon. That's why "I can build it" is no longer an argument. Most ideas out there don't solve any pain.
Key concepts
Real demand, technical excitement, pain, a market that "screams," builder's idea.
What it is
Build from someone's concrete pain instead of building just for the sake of solving something.
Why learn it
Adopting a new product takes work: signing up, entering a credit card, learning how to use it. Only people in enough pain will pay that switching cost.
Key concepts
Pain, switching cost, urgency, intensity, frequency.
What it is
Project the idea into the near future and ask whether it will still have value as models and tools advance.
Why learn it
What requires a product today could become a free feature, or a single request to a model, in a few months. Building this means building something with a short shelf life.
Key concepts
Shelf life, "vibe-codeable," one prompt, feature vs. product, 3/6/12-month horizon.
What it is
An example of a niche chosen deliberately because it's unappealing and constantly renews itself: protecting apps built with AI.
Why learn it
It puts the earlier criteria into practice: real pain, a problem that can't be solved once and for all, and one that grows alongside the technology.
Key concepts
Cybersecurity, cat and mouse, rising attacks, open models without safeguards, defense with frontier models.
What it is
Many pain points aren't about lacking a tool, but lacking knowledge: someone knows a risk exists but doesn't know how to assess it.
Why learn it
This kind of pain is acute because it comes with uncertainty. And people don't want to spend time or tokens on it, even when they know they could.
Key concepts
Information gap, uncertainty, hiring a specialist, time and tokens, outsourcing the worry.
What it is
Prefer unglamorous problems that few people want to solve over flashy problems everyone is tackling at the same time.
Why learn it
“Sexy” problems make for good demos, so they get crowded. And some popular niches are already well served by AI chat itself, with no SaaS needed.
Key concepts
Sexy problem, boring problem, saturation, content creation, intimate process.
🐘 Giants, data, and defensibility
In a world where any software can be copied, the question isn’t “Do I have a moat?” but “What, exactly, makes me hard to copy, and what’s left if someone does?”
Module topics
What it is
Assess whether a large company with plenty of resources could wake up tomorrow and launch your product, and whether it would have the credibility in that market to take your customers.
Why learn it
Being copied by a giant is the biggest risk for a small product. But not every giant has the right reputation in every niche.
Key concepts
Cloning, credibility in the niche, independent product, frictionless experience, landing page designed for the customer.
What it is
Imagine your product as a plugin, app, or feature inside ChatGPT, Claude, or Gemini, accessible with one click and no new subscription.
Why learn it
If an 80% good version shows up inside something your customer already pays for, they don’t have to discover you, trust you, or enter their credit card. That’s the concrete risk.
Key concepts
Plugin store, 80% of the way there, zero adoption cost, “Is it worth a prompt for them?”, throw them a bone.
What it is
Ask whether, even if the product is surpassed, the data it generated has value of its own, enough for someone to buy it.
Why learn it
Companies have sold for fortunes, not because of their software or users, but because of the dataset only they had.
Key concepts
Intrinsic asset, dataset, acqui-hire, hard-to-generate data, data buyers.
What it is
Plan from the start what data product usage will generate, how to store it, and how it could become valuable.
Why learn it
Good data doesn’t appear by chance. If you don’t design for it from the start, you’ll lose the history that could make the business valuable even in the worst-case scenario.
Key concepts
Data as a byproduct, anonymization, structure, consent, plan B.
What it is
Treat ease of building as a warning sign: if the product is trivial to make, it’s trivial to copy.
Why learn it
In a world where anyone can build anything, difficulty has become a form of protection.
Key concepts
Difficulty as a moat, one-shot, iteration, rare epiphany, accumulated effort.
What it is
The real value is in the invisible work that makes a result feel easy and immediate to the user.
Why learn it
The user sees a text box and a button. A competitor copies the box and button. What they can’t copy is the pipeline of decisions between input and output.
Key concepts
Proprietary pipeline, nuance, end-to-end design, instant experience, care.
🏒 Use, recurrence, and habit
A product only matters if people keep using it. This module covers anticipating the market, choosing where the window stays open longest, and becoming part of the customer’s routine.
Module topics
What it is
The hockey analogy: don’t skate to where the puck is; skate to where it’s going. For a product, that means building for the market as it will be a few months from now.
Why learn it
If you build for the current state, you arrive after the opportunity has passed. The pace of change in AI makes that window much shorter.
Key concepts
Anticipation, cannibalization, window of opportunity, extrapolating trends.
What it is
The window of time when your audience hasn’t yet realized they could build on their own, with AI, what you’re selling.
Why learn it
Even if the solution can be replicated, many customers won’t realize it for 6 to 12 months. That time is an opportunity, as long as you have a plan for what comes next.
Key concepts
Arbitrage, market naivety, 6-12 months, scale, sell, or pivot.
What it is
Sell to traditional sectors that make a lot of money doing one thing well and don’t keep up with AI advances.
Why learn it
In these sectors, arbitrage on customer naivety lasts much longer. Even with far more capable models, they’ll still be learning the basics.
Key concepts
Slow-moving industry, oil and gas, metals, physical world, slow adoption cycle.
What it is
Distinguish problems that are never fully solved (and support a subscription) from problems that can be solved well enough and become saturated.
Why learn it
Customer retention and lifetime value depend on this. A saturated problem becomes free; a problem that keeps renewing itself supports ongoing charges.
Key concepts
Recurring revenue, customer lifetime value, saturation, open source, “Why would I pay?”
What it is
Make your product part of the customer’s routine, working in the background and delivering progress they couldn’t make on their own.
Why learn it
The biggest factor in keeping a customer is being part of their workflow. Products that rely on users remembering to open them get forgotten.
Key concepts
Daily workflow, background, proactive, synergy with automation, passive progress, data the user already has.
What it is
Two other signs of recurring use: the product creates a pleasant feeling (relief, security) and has the potential to become an everyday verb.
Why learn it
A good feeling makes users come back on their own. Becoming a verb is rare, but when it happens, the language itself does the marketing.
Key concepts
Feeling of security, relief, antivirus, manufactured demand (ethical warning), brand as a verb.