Nine Laws of Retention (15 Years of Data)

1. Bad retention is un-fixable.
   Push alerts, drip e-mails and button-colour A/B tests will not turn 10 % D1 retention into 40 %.

2. Retention only falls; it never rises.
   The curve decays in a predictable half-life—D7 ≈ 50 % of D1, D30 ≈ 50 % of D7.

3. Revenue retention can grow even while user retention shrinks.
   Slack loses seats but gains dollars; Amazon starts with books, ends with everything.

4. Category is destiny.
   No-one opens a hotel-booking app daily—accept the cadence or pick another vertical.

5. Scale dilutes quality.
   Early “golden” users arrive organically; later paid cohorts convert and retain worse.

6. Churn is asymmetric.
   Losing a user is trivial; winning them back is nearly impossible.

7. Retention is noisy.
   Seasonality, bugs and new Android builds make month-to-month comparisons almost meaningless.

8. Viral + low retention = crash.
   TikTok-fuelled spikes feel great until the music stops and you own a ghost town.

9. Great retention feels like magic.
   When you see 50 % D30, you know the team hit a breakthrough insight, not a local-maximum tweak.

The 10 % Rule: When to Pivot
If D1 retention is < 10 %, stop optimising. Strip the product to the studs or change the category.
Real-world playbook:

• Re-wire the home-screen (feed → step-by-step wizard).

• Swap the core verb (share → collect).

• Re-position against an incumbent (“We are the Notion of X”).
  The bigger the swing, the higher the survival rate.

Retention Is a Category Choice
Want daily habit? Build where daily habits already exist—chat, search, social, music, news.
Enter with 20 % differentiated wedge (stories vs. posts, cursor vs. IDE, infinite vs. paginated) powered by a new enabling layer (GPU-accelerated filters, LLM copilots, 5G video).
Timing > novelty. Ask “Why now?”—if you can’t answer, you’re just rearranging deck chairs.

The Myth of the “New Market”
Most smash hits are late entrants: Google (10th search engine), Instagram (Hipstamatic + feed), Tesla (EV #N).
They rode existing demand with a 10× better wedge and cleaner business model.
True category creation—Uber, ChatGPT—is rare and expensive. Unless you have five years and a research lab, compete where the users already are.

Bottom Line
Stop polishing bricks—move to a new foundation. Pick a high-frequency arena, attack an incumbent with a 20 % breakthrough, and ship before the window closes. Retention is not a feature; it is the side-effect of being in the right category at the right time with the right twist.

Nine Laws of Retention (15 Years of Data)

1. Bad retention is un-fixable.
   Push alerts, drip e-mails and button-colour A/B tests will not turn 10 % D1 retention into 40 %.

2. Retention only falls; it never rises.
   The curve decays in a predictable half-life—D7 ≈ 50 % of D1, D30 ≈ 50 % of D7.

3. Revenue retention can grow even while user retention shrinks.
   Slack loses seats but gains dollars; Amazon starts with books, ends with everything.

4. Category is destiny.
   No-one opens a hotel-booking app daily—accept the cadence or pick another vertical.

5. Scale dilutes quality.
   Early “golden” users arrive organically; later paid cohorts convert and retain worse.

6. Churn is asymmetric.
   Losing a user is trivial; winning them back is nearly impossible.

7. Retention is noisy.
   Seasonality, bugs and new Android builds make month-to-month comparisons almost meaningless.

8. Viral + low retention = crash.
   TikTok-fuelled spikes feel great until the music stops and you own a ghost town.

9. Great retention feels like magic.
   When you see 50 % D30, you know the team hit a breakthrough insight, not a local-maximum tweak.

The 10 % Rule: When to Pivot
If D1 retention is < 10 %, stop optimising. Strip the product to the studs or change the category.
Real-world playbook:

• Re-wire the home-screen (feed → step-by-step wizard).

• Swap the core verb (share → collect).

• Re-position against an incumbent (“We are the Notion of X”).
  The bigger the swing, the higher the survival rate.

Retention Is a Category Choice
Want daily habit? Build where daily habits already exist—chat, search, social, music, news.
Enter with 20 % differentiated wedge (stories vs. posts, cursor vs. IDE, infinite vs. paginated) powered by a new enabling layer (GPU-accelerated filters, LLM copilots, 5G video).
Timing > novelty. Ask “Why now?”—if you can’t answer, you’re just rearranging deck chairs.

The Myth of the “New Market”
Most smash hits are late entrants: Google (10th search engine), Instagram (Hipstamatic + feed), Tesla (EV #N).
They rode existing demand with a 10× better wedge and cleaner business model.
True category creation—Uber, ChatGPT—is rare and expensive. Unless you have five years and a research lab, compete where the users already are.

Bottom Line
Stop polishing bricks—move to a new foundation. Pick a high-frequency arena, attack an incumbent with a 20 % breakthrough, and ship before the window closes. Retention is not a feature; it is the side-effect of being in the right category at the right time with the right twist.