The ten steps for a successful behavioural science intervention

Behavioural science becomes expensive decoration when teams sprint to step seven. The intervention is the sexy bit. Defaults! Norms! A message with a psychologically informed adjective! Everyone can point at something launched. Mapping the behaviour, defining the outcome and admitting the problem may have several causes feels slower and dangerously like thinking.

So a campaign appears before anyone has agreed what success means. Awareness rises, according to a survey written by the people who raised it. The target behaviour remains politely unchanged. The ten steps are valuable because they make intervention design look like the middle of the job. Scope comes first. Behaviour must be specific enough to observe. Causes need separating: inability, inconvenience, uncertainty, habit and social risk will not all surrender to the same clever email. Implementation then meets reality, measurement follows, and the result is allowed to alter the next action.

This sounds obvious. Most useful disciplines do. The failure comes from treating a list as compliance: ten boxes ticked, no uncomfortable decision made. “Assess effects” means little if the team has already promised success and booked the awards entry. A nudge is one possible response to a diagnosed barrier inside a system, with no magical privileges. If step seven has been chosen before steps one to six, the science is being used as garnish—and the garnish has started invoicing by the day.

Behavioural principles

Behavioural ideas at play in this post

Short, plain-English explanations of the principles behind this post, with links to related books and examples in the archive.