Think Week

Part V · Toolkit · Chapter 17 · 2 min read

The one-page memo

Every formula and test from the course on one page, plus the decisions this Think Week produced and a template for your own.

In this chapter
  1. All the formulas
  2. All the tests
  3. What this Think Week decided
  4. Your own one page

All the formulas#

Formula Chapter
Your limit is your 4,000 weeks, not your capability 1
A Think Week = separation + a question + input + a written output 2
Judge the process, not the outcome 3
Give the next hour to the highest marginal value; set floors, then everything to success 4
When X gets cheap, its complements get expensive 5
Interfere where the machine can't check and much depends on it 6
A segment is real only if you can run the query 7
Luck surface ≈ doing × telling 7
Kill criteria before you start 8
Success ≈ Demand × Hole × Access × Edge × Durability 9
Service → niche product → publish what you learn 10
Sell → deliver by hand → automate 10
Stable = existing people × unchanged need × accountable × tool-independent × monthly 11
Sell the slow layers; the fast layer is your cost 11
New code, old customers 13
Revenue = market size × your share 13
Brands out front, one engine and one audience underneath 14
Stage gates decide, not the calendar 14

All the tests#

Four kill filters (Chapter 8), cheapest first:

  1. Are people already spending money on this?
  2. Can I reach them at this price, within my budget?
  3. Will I get a real signal within 2 months?
  4. What do I keep if it fails?

Five durability questions (Chapter 10):

  1. Is the value in something AI can't produce?
  2. Does it get better when models get better?
  3. Do I own the customer relationship?
  4. Would leaving me hurt the customer?
  5. Has the need lasted 20 years, and will it last 20 more?

Before any bet (Chapter 8): max weeks · continue signal at week 6 · stop signal · what I keep.

What this Think Week decided#

  1. Hold one audience: small independent publishers who want to own their readers. Hold the people, not the platform.
  2. Owldraft is the umbrella, a media and publishing house. Each product is an independent brand with one mission; Owldraft sits in the footer.
  3. First brand: a newsletter inside WordPress (Buttondown-style), with hosted sending as the business model. Target: 10 paying customers by week 8, or change the angle.
  4. One shared engine and one shared audience under every brand. Build the engine cleanly from day one.
  5. Stage gates decide when the next brand starts: publication brand, then local news, then let the market decide.

Stopping:

  • Fixing the old marketplace shop's reputation. The flagship theme goes to maintenance mode and its customers become a bridge.
  • Trying to make one product serve every mission.
  • TV and scrolling. (Sleep, exercise, friends and one pointless hour a week stay as floors.)

First step for Monday:

  • Count the past customers who can actually be emailed.
  • Run one technology lookup for WordPress sites using a paid newsletter tool.
  • Write the kill criteria for the newsletter brand before sending a single email.

Your own one page#

Template

The question I brought:

3–5 decisions:

What I'm stopping (my "other 20"):

The bet, with kill criteria: max weeks · continue if by week 6 · stop if · I keep

Floors I won't break: sleep · body · people · play

First step for Monday:

Next mini Think Week (in 12 weeks):

Analysis decides where you stand. The world decides what happens to you there. So stand close to what you know, make lots of cheap contact with people, and spend your weeks on what AI can't make cheap.