The Behaviour Thinking process.
Six moves for taking one behavioural bet from fog to verdict.
The rhythm
Behaviour Thinking moves through a simple rhythm. Name the behaviour. Understand the current route. Find the collision. Test the conditions. Weigh exposure against evidence. Decide what to change.
The six moves are taught in order because order helps people learn. In practice, teams loop back. A better current-route insight may rewrite the bet. A collision may reveal the wrong actor. Evidence may force the team to simplify.
The worked example
A retailer is rolling out an AI stock-planning tool to sixty stores. The pilot worked. The business case is strong. The risk is whether store managers act on the recommendation every Monday morning. This one example runs through all six moves below.
The six moves
Name the bet
What must someone actually do?
Write the behaviour the plan depends on as one structured sentence. Keep it filmable: if you could not point a camera at it, it is not a behaviour yet, it is an outcome wearing a costume.
For this to work, [actor] must [behaviour] at [moment], despite [pressure].
"Improve stock accuracy" is an outcome. "Store managers review and act on the tool's recommendations every Monday morning, despite staffing gaps" is a behaviour.
Respect the current route
What do people do now instead, and why does it make sense?
Describe the existing behaviour with respect. The old route is your real competitor, and it is undefeated at home. Plans that treat it as laziness or resistance design interventions for a problem that does not exist.
The managers trust their own judgement. It is faster, it has never failed them publicly, and knowing your store without needing a system is a point of pride.
Find the collision
Where does the new behaviour meet what the current system rewards, protects or normalises?
The collision is where the plan will actually be decided. Everything else is upstream of it.
The organisation praises decisive local judgement and firefighting, and the tool asks managers to defer to a recommendation. Monday morning is also the week's worst moment for reflective work.
Test the conditions
What pulls people back, what supports action, and what is missing?
Ask what the behaviour asks of people, and what would have to hold it up: training, yes, but also who owns it, which ritual carries it, what a manager's manager asks about on Wednesdays. Behaviour that has to be sustained by willpower alone is already failing, just slowly.
The Monday behaviour has no owner, no ritual and nothing in the weekly rhythm that asks whether it happened. Support so far is a training session and a login.
Weigh exposure against evidence
What breaks if the behaviour does not happen, and what proof do we have that it will?
Danger zone
High exposure. Weak evidence. Exactly where organisations most often scale.
Protected
High exposure, strong evidence. Proceed with monitoring.
Cheap to test
Low exposure, weak evidence. Test before you spend more.
Safe ground
Low exposure, strong evidence. Get on with it.
Exposure is high: the whole model rides on Mondays. Evidence is weak: the pilot may have worked because someone at head office phoned every store each Monday, which is enthusiasm doing the job the design was supposed to do.
Decide
What verdict should the work receive?
The retailer's verdict: test the Monday behaviour properly in six stores before scaling, and roll the tool out as a weekly operating behaviour rather than a technology installation.
The decision is the product.
What the process produces
The output is not a report. The output is a better decision, built from a named behavioural bet, an honest reading of the current route, the collision, a condition map, an exposure and evidence judgement, a verdict, and a next move.
- "We expect strong adoption"
- "Engagement will improve"
- "The culture will shift"
- A named, filmable behaviour
- An honest exposure/evidence rating
- A decision the team can act on