Think Week

Part III · Choosing a bet · Chapter 8 · 4 min read

Timing, and ten more factors

The right moment can make a bet 2–20× easier. Here's what's hidden inside "timing", and the other dimensions that change the game.

In this chapter
  1. New market vs mature market
  2. What's hidden inside timing
  3. Ten more factors
  4. Filter like a database
  5. Decide when you'll quit, before you start

New market vs mature market#

The first split is simple:

New market (e.g. AI agent plugins) Mature market (e.g. WordPress themes)
Rules Easy, loose Hard, strict review
Competition Little Heavy
Education You may have to teach people it exists Nobody needs educating
Nature A bet An investment with data: known target to beat

A mature market is like venture investing with full information. What you invest isn't money; it's your effort.

What's hidden inside timing#

  1. Data says timing matters most. Bill Gross (Idealab) studied ~200 startups. Timing explained 42% of the difference between success and failure. Team 32%, idea 28%, business model 24%, funding 14%.
  2. There's a third state: a mature market going through a shift. Demand is proven, so nobody needs educating. But the rules just reset (because of AI), and the big players are slow, because changing hurts the business they already have (Christensen's innovator's dilemma). This is often the best spot of all. Carlota Perez makes a related point: after the frenzy of building a new technology, lasting money comes from applying it to old, boring industries.
  3. The chasm. Buyers arrive in waves: early adopters buy the vision and forgive rough edges; the majority wants proof and references. Many products die in the gap (Geoffrey Moore, Crossing the Chasm). Know which group you're selling to.
  4. Platforms are generous early, then take back. A new marketplace gives free visibility, easy review and few competitors, because it needs an ecosystem. Later: crowded, ranking games, fees, and maybe the platform builds your feature itself. Get in early, but move your audience onto your own list before the doors close.
  5. Let someone else pay to educate. "A new market needs education" is only a problem if you pay for it. When a big company promotes a new category, ride that wave. Rule: enter a new market only where a bigger player pays for education.
  6. Too early looks the same as wrong. A sign the market is ready: people already hack around the problem, with manual workarounds and forum posts asking "how do I…".
  7. The AI clock runs fast. WordPress themes took ~10 years from new to mature. AI markets move through the same phases in 1–2 years. A 2-year bet is very long in AI time, so prefer bets you can test in weeks.

Ten more factors#

# Factor The question Why it matters
1 Existing spend Are people already paying for a solution, even a bad one? Money already flowing = no education needed, price known. The strongest single signal.
2 Painkiller or vitamin Does it hurt now? Pain sells itself. "Nice to have" needs marketing forever.
3 Price decides the channel Can this price pay for this query? $29 can't pay for sales calls; $2,000 can. $1M/year = 1,000 × $1,000 or 100,000 × $10: completely different games. Choose price and channel together.
4 Owned vs rented distribution Who controls the audience? Your list and past customers are owned. Marketplaces, Google, social are rented: the landlord can change rules overnight. Past buyers are your cheapest query.
5 Does AI progress help or kill it? When the next model ships, am I happy or scared? If you patch today's model weakness, the next release kills you. If you get better as models improve, each release is a gift.
6 Time to signal How fast will I know? Marketplace launch: 2–4 weeks. SEO: 6–12 months. Enterprise sales: 6–9 months. Fast feedback prevents wasted years.
7 Compounding Does each week stack? Content, audience, reviews, reusable code, reputation stack. Client projects, paid ads, one-off launches reset to zero.
8 What's left if it fails Do I still own something? An audience, skill, codebase or reputation means the year wasn't wasted (Taleb's asymmetric bets).
9 One winner or many Does the 50th player still earn a living? Solo builders should prefer markets with room for many, and markets too small for funded companies to care about.
10 One-time or recurring Do I start from zero every month? Recurring revenue stacks.

Two more worth checking: built-in visibility (a theme shows on every site that uses it; a backend plugin is invisible), and earned secrets (what you know from years inside a market that outsiders can't see).

Filter like a database#

A database runs the cheapest filters that remove the most rows first. Do the same with ideas:

Decide when you'll quit, before you start#

From Annie Duke's Quit. Fill this in before any bet:

  • The most weeks I'll spend: __
  • The week-6 signal that means continue (e.g. 20 paying customers, not likes): __
  • The signal that means stop: __
  • What I keep if I stop: __

Kill criteria

Deciding when you'll quit, before you start, is what stops a 6-month bet from becoming 2 years.

Takeaways

  • Timing explains more than the idea does. Look for the third state: proven demand, reset rules.
  • Enter new platforms early, but move the audience to your own list.
  • Existing spend is the strongest signal. Price and channel are one decision.
  • Run cheap filters first. Write kill criteria before you start.