Sweet spots: 12 moves

The master formula#

You're never the force that moves the market. You surf its waves. With 10–20 factors in play, choosing a sweet spot can make things 2–20× easier, and choosing wrong can make them impossible. What you're looking for is a market hole.

• Demand: people already want this, ideally already pay for it.
• Hole: the gap between what they want and what exists.
• Access: you can reach them cheaply (your query is affordable).
• Edge: why you: skill, audience, speed, or knowledge others lack.
• Durability: how long until the hole closes (competitors, the platform, AI).

The factors multiply. If any one is zero, the result is zero. Try it:

How to make the huge space searchable: start from a proven success (demand already proven) and apply one move. 50 successes you know × 12 moves = 600 candidates, and every one of them starts with proven demand.

In the tables, the AI era column means: born = started in the AI era, holds = still fine or growing, hurt, broken.

1. Subtract: unbundle a big flexible tool#

flexible tool − options + one segment's whole job = new product

The limits are the product. WordPress has too much flexibility; a good product needs constraints.

2. Narrow: same product, one segment's words#

product for everyone → same product for one segment (their words, their channels)

No new code, only a new answer to "who is it for?" This is the "copy it, put it in the right place" hole.

3. Move: same thing, new place#

proven in place A → copied to place B (country, language, platform)

Local advantages (payments, laws, support, relationships) survive AI. Pure translation doesn't.

4. Cross: X + Y#

a segment needs X and Y; X-experts ignore Y, Y-experts ignore X

WordPress × a niche survives because it lives inside the customer's site, with their data and daily work.

5. Package: raw tech → one job#

raw capability + one job + one segment = product; the user never sees the tech

Wins on speed and distribution; dies when the platform ships the feature. Take the cash, or turn it into move 8 or 11.

6. Invert: do what the leader can't#

the leader earns from M; doing X would hurt M → X is safe for you

The best kind of hole: the leader chooses not to fill it. It stays open as long as their business model does.

7. Rebuild on a shift#

proven market + a shift (new tech, ranking rule, behavior) → rebuild before the leaders move

The "third timing state" from Chapter 8: proven demand, a slow leader.

8. Own the query: audience first, product second#

free value to one niche → an audience you own → learn their pains → sell with almost no ad cost

An audience that finds you through Google is rented, and AI cut that path. Tailwind's docs traffic fell 40% while the framework got more popular. An audience that knows your name, face or inbox is owned, and survives.

9. Sell shovels#

a wave creates many builders → sell what all of them need

Sell to AI builders, not against AI.

10. Go boring or small#

unexciting but essential job + tech people ignore it → little competition, loyal customers

"Too small for the big players" protects a solo person.

11. Productize a service: sell the outcome#

a service done by hand → fixed scope and price → then software/AI does most of the work

Customers want "done", not a tool. When AI does the work, AI is your cost, not your competitor.

12. Arbitrage: buy cheap here, sell dear there#

The gap always closes. Fine for quick cash, bad for a 4,000-week plan.

What the AI era broke#

What broke: selling information, templates, generic content or thin wrappers, and relying on search traffic.

What survives:

1. Trust, and an audience you own.
2. Living inside the customer's daily work and data.
3. Products that get better when AI models get better.
4. Selling shovels to AI builders.
5. Holes the leader structurally can't fill.

Sources: Tailwind layoffs · Base44 / Wix · Gamma · Cal AI · Surge AI · PhotoAI · Lovable · Midjourney · Chegg · Tally · n8n · Supabase · ConvertKit · HeadshotPro · ThemeForest sales drop