Behavioural Change Toolkit · Deliver

Workshop — Design the Nudges (2 hr)

Purpose: turn diagnosed barriers into a designed, filtered intervention stack. Who: 6–8. Must include someone who does the work, someone who can change the system, and someone who controls a budget or a process. Prerequisite: Workshop — Behavioural Diagnosis (90 min) or an equivalent evidenced diagnosis.

⚠️ Don’t run this without the diagnosis. A room generating ideas against an unevidenced barrier produces a confident plan aimed at the wrong thing.

Run sheet

0:00 — Recap the diagnosis (10 min)

Behaviours and barrier scores on the wall. Read the evidence, not just the scores. One quote each. The quotes stop the room drifting back to assumption.

0:10 — The constraint (5 min)

“Rule for today: every idea has to name the barrier it removes. If it doesn’t, it goes on the parking wall. We’re also going to be suspicious of any idea that’s an email.”

0:15 — EAST sprint, round 1: EASY (20 min)

“How could we make this so easy it’s almost automatic?”

Silent write 5 min → share 10 min → build 5 min.

Prompts to inject when it slows:

What’s the default right now, and can we flip it?
How many clicks? Which three can we delete?
What could be pre-filled?
Can we switch off the old route entirely?
What would we do if we had to make the wrong behaviour harder than the right one?

0:35 — Round 2: ATTRACTIVE (15 min)

“What’s in it for the person doing it — this week, not next quarter?”

Push hard on immediacy. Anything delayed beyond a fortnight is weak.

0:50 — Break (10 min)

“Who do these people actually listen to? How do we make the right thing visibly normal?”

Prompts: who’s already doing it well? Who’s the credible messenger — not the senior one? What can we make visible that’s currently invisible? What are leaders doing that contradicts this?

1:15 — Round 4: TIMELY (10 min)

“What’s the exact moment this behaviour has to happen? What already reliably happens right then?”

Output: a cue for each behaviour. See Habit Loop and Behavioural Drift.

1:25 — APEASE filter (20 min)

You’ll have 25–40 ideas. Cluster duplicates → around 12.

Score together on a flipchart grid, 1–5: Acceptability · Practicality · Effectiveness · Affordability · Side-effects · Equity.

Facilitator moves:

Go fast. First instinct, no debating single points.
When a senior person’s pet idea scores badly, let the framework say it. Don’t rescue it and don’t gloat.
Force the equity question: “Does this work on night shift? For part-timers? For the site with the old kit?” It’s routinely missed and expensive to discover later.

1:45 — Build the stack (10 min)

Select 3–4 across different layers:

Layer	Selected
Make it easy (environment)
Make it normal (social)
Make it cued (habit)
Make it rewarded
Make it required

⭐ At least one must be an environment change, not a communication. If the whole stack is messages, the room has designed a comms plan and called it an intervention.

1:55 — Owners and dates (5 min)

Every selected item: named owner (a person, not a team) and a date. Unowned items don’t happen, and an item owned by “the project” is unowned.

Output

Intervention	Barrier removed	Owner	By	APEASE total

→ T — Intervention Note each · test with Pilot Design and Experiment Card · into Change on a Page

Common failure modes in this session

Symptom	Fix
Everything generated is a communication	“What would work if nobody read their email?”
Ideas are all training	Point at the barrier scores. Training only fixes Capability.
Room designs for the ideal user	“What about someone on their fourth night shift?”
Senior person’s idea dominates	Silent scoring, revealed simultaneously
Nothing is chosen	You’ve run out of time on generation. Hard-stop the sprints next time.

← Back to the Behavioural Change Toolkit

Last updated: 11 August 2026