# Improve Retention

> Behavior design for user retention using B=MAP — an AI agent skill based on *Tiny Habits* by BJ Fogg.

Published 2026-02-19 · Updated 2026-08-17 · Product & Innovation · Free and MIT-licensed
Canonical: https://skills.wondel.ai/skills/improve-retention/
Source: https://github.com/wondelai/skills/tree/main/improve-retention

**Install:** `npx skills add wondelai/skills/improve-retention --global`

## What is the Improve Retention skill?

Improve Retention packages BJ Fogg’s Tiny Habits, and the B=MAP model behind it, as a skill your agent loads when usage is sliding. Every behavior that failed to happen gets blamed on exactly one of three things — motivation, ability or prompt — and the fix is usually to shrink the action until it survives a bad day.

## Key concepts from Tiny Habits your agent applies

- **B=MAP Model** — Behavior happens when Motivation, Ability, and Prompt converge at the same moment.
- **Shrink the Behavior** — Make the target action so small it's impossible to fail — then celebrate the tiny win.
- **Ability Chain** — Identify the weakest link: time, money, physical effort, mental effort, or routine disruption.
- **Prompt Design** — Place prompts where motivation and ability are highest — after an existing habit or key moment.
- **Celebration & Shine** — Create an immediate positive emotion after the behavior to wire the habit into the brain.

## When to use Improve Retention — and when not to

Reach for it when:

- Activation looks fine and week-two retention does not: people complete the core action once, then quietly stop.
- Your reminder emails go out at nine in the morning to everyone, whether or not that is a moment anyone could act.
- The core action takes eleven minutes and a decision, and you keep trying to solve that with more encouragement.
- You are about to ship a streak counter without knowing whether the blocker is effort, timing or wanting to at all.

Reach for something else when:

- People leave because the product never did the job they hired it for — restart with Jobs to Be Done.
- You want a compounding engagement cycle with variable rewards and investment, not one behavior made easier — use Hooked UX.
- The action fails because the screen is confusing rather than effortful — UX Heuristics finds usability faults faster.
- You need to know which retention number to watch and what counts as good at your stage — that is Lean Analytics.

## About BJ Fogg

Founder of Stanford Behavior Design Lab. BJ Fogg is a behavior scientist and founder of Stanford's Behavior Design Lab. His research has influenced the design of products used by billions, and he has trained innovators at companies like Instagram, Fitbit, and Uber.

More: https://en.wikipedia.org/wiki/B._J._Fogg · https://www.wikidata.org/wiki/Q4834030
The book: https://openlibrary.org/books/OL27906569M/Tiny_Habits

## Example prompts

- Analyze why users drop off after signup and design habit loops using improve-retention skill _(Retention analysis)_
- Apply B=MAP to improve daily active usage of our app using improve-retention skill _(Engagement)_
- Design prompts that catch users at their highest motivation using improve-retention skill _(Notification design)_
- Reduce friction in our core workflow using the Ability Chain using improve-retention skill _(UX optimization)_

## Frequently asked questions

### How do I install the Improve Retention skill?

Run `npx skills add wondelai/skills/improve-retention --global`. It takes about 30 seconds and needs no account. The Improve Retention skill then works in Claude, Claude Code, Claude Cowork, Codex, Cursor, OpenClaw and Hermes Agent — anything that reads the open agentskills.io format — and your agent loads it on its own when a task calls for it. It is free and MIT-licensed, and the source is at https://github.com/wondelai/skills.

### Which book is the Improve Retention skill based on?

It packages *Tiny Habits* by BJ Fogg — founder of Stanford Behavior Design Lab. The skill distils the book's method into instructions your agent follows while it works, covering b=map model, shrink the behavior and ability chain. It sits in the Product & Innovation part of the library.

### Isn’t this the same as the Hooked UX skill?

No, and the difference is diagnostic. B=MAP says a behavior occurs only when motivation, ability and a prompt converge at the same moment — so when it does not occur, precisely one of three things is missing and you go and fix that one. It is a model for behaviors that are failing, including ones that have never happened. Hooked starts from a behavior that already happens and engineers its recurrence. Different questions, and answering them in the wrong order is why so many retention projects ship a streak nobody wanted.

### Why shrink the behavior instead of motivating people harder?

Because motivation is the least reliable of the three inputs. It spikes and crashes on a schedule you do not control, so any plan that requires a motivated user is a plan that works on good days only. Ability you can engineer permanently. Fogg’s Ability Chain names five links — time, money, physical effort, mental effort, and disruption to routine — and only the weakest one matters, so there is normally a single thing to fix rather than five. Shrink until the action survives a distracted Tuesday.

### Tiny Habits is about personal habits — does it translate to a product?

Mostly, and the translation is what the skill does. The book’s examples are flossing and push-ups, but the mechanics port cleanly: a prompt is a notification or an existing screen, ability is your flow’s cost in time and thought, and the anchor — *after I do X, I will do Y* — becomes attaching your action to something the user already does. The element teams drop is celebration. Fogg argues immediate positive emotion is what wires the habit, which in software means a fast, legible confirmation rather than a silent toast.

### Will this fix churn?

Only the behavioral half of it. It works on users who intend to come back and do not — forgetting, friction, bad timing — and does nothing for users who have decided the product is not worth returning to. Read your exit surveys first: “I forgot” and “too much hassle” are behavior design problems, “I did not need it” is a demand problem, and treating the second as the first buys you one month and costs you a quarter.

## Related

- All Product & Innovation skills: https://skills.wondel.ai/skills/product/
- Jobs to Be Done: https://skills.wondel.ai/skills/jobs-to-be-done/
- Hooked UX: https://skills.wondel.ai/skills/hooked-ux/
- Lean Startup: https://skills.wondel.ai/skills/lean-startup/
- Design Sprint: https://skills.wondel.ai/skills/design-sprint/
- Continuous Discovery: https://skills.wondel.ai/skills/continuous-discovery/
- Inspired: https://skills.wondel.ai/skills/inspired-product/
- The Mom Test: https://skills.wondel.ai/skills/mom-test/
- The 37signals Way: https://skills.wondel.ai/skills/37signals-way/
- Lean Analytics: https://skills.wondel.ai/skills/lean-analytics/