Tool: Behaviour Decay Canvas

Lauren Kelly

A practical map for catching the moment a new behaviour starts to slide

Purpose: Spot, explain and fix the drop that usually follows a big launch.
Who uses it: Product teams, service designers, marketers and change leads who need habits and longer term retention.
When to pull it out: During planning, two weeks after launch (if you missed it during planning), and any time adoption stalls.

Behaviour goal. The action you need people to repeat.
Six week timeline. Week 0 to Week 6 in columns.
Likely blockers. One row for each friction, motivation dip, or context clash you can foresee. Use COM-B or BMAP here for extra behavioural insight.
Reinforcement ideas. A matching row where you list nudges, tools or rituals to counter each blocker.
Signals to watch. Metrics or qualitative cues that show the blocker is real or the fix is working.

Print it big or drop it into Miro. One behaviour per canvas keeps focus sharp.

Define the behaviour goal
Write the repeat action in a sentence. Example: "Managers log coaching notes in the system every Friday."
Sketch the timeline
Draw seven vertical blocks: Pre‑launch, Week 1, Week 2 … Week 6.
Brainstorm blockers
In each week, ask the team:
- What motivation might fade here?
- What context friction might return?
- What competing priorities could appear?
  Stick each blocker onto the right week.
Design reinforcement
For every blocker add at least one counter‑move. Think prompts, social proof, shortcuts, new rewards. Make them specific and timed.
Choose signals
Decide how you will know the blocker hit or the fix worked. Use a metric (daily actives) or a qualitative cue (support tickets about feature X).
Assign owners and dates
Put names and deadlines beside each reinforcement. No owner means no action.
Review weekly
At the end of every week, mark blockers that showed up and reinforcements that fired. Adjust the next week’s plan before Monday morning.

Balance motivation and friction. Remove effort and add reward in the same window.
Start with evidence. Past rollouts reveal common dips. Use them.
Keep reinforcements small. A quick prompt or story often beats a full relaunch.
Watch real users. Shadow two people in week two. Their pain points beat any guesswork.
Pair quantitative with qualitative. Numbers tell you when, conversations tell you why.
Plan for ownership changes. Projects hand over. Build duties into normal roles early.
Translation strategy. A behavioural signal is not the same as a service or product signal. Think, where are we already collecting data to show if behaviour x is turning up?