Developed by BJ Fogg (A Behavior Model for Persuasive Design, 2009).
Holds that a behavior happens when motivation, ability, and a prompt meet at the same moment, so the dependable fixes are making the action easier and the prompt reliable.
One salesperson, one expense entry: shrink the task or raise the motivation
The square is one salesperson on one day: how easy logging an expense is, and how much she wants to. The curve is the action line: above it, the prompt (a buzz when her card is used) gets the entry logged; below it, the prompt is ignored. Move both sliders and see which gets her over the line on an ordinary day. Illustrative.
When the card buzzesAbove the line: the buzz arrives and she logs the expense.Below the line: the buzz arrives and she swipes it away.
An example
A field sales team of forty was supposed to log every expense on the day it was spent. About half did. The operations manager tried motivation first: a note from the sales director, then a leaderboard of the most current filers. Same-day logging rose for two weeks and slid back.
Then she timed the task. Logging one client lunch took eleven to fourteen minutes: open a laptop, sign in through the VPN, pick a cost center from a list of ninety, type the amount, scan the receipt. She cut it to one photo. The expense app now opens on the camera, reads the amount, and guesses the cost center from the merchant. She also fixed the prompt: the phone buzzes the moment the corporate card is used, while the receipt is still in hand. Within a month same-day logging went from 52 to 88 percent, and nobody's motivation had changed.
Why it works
Motivation and ability trade against each other. A hard task gets done only on a high-motivation day; an easy one gets done even on a low day. Of the three, motivation is the one to count on least: it moves with mood, workload, and the hour, and a campaign to raise it fades within weeks. Ability is in the designer's hands, and once a task is made easier it stays easier.
The prompt is the piece most easily overlooked: a person who is willing and able, and never cued at the right moment, does nothing. So the model turns a vague complaint (people will not do it) into three questions with different fixes, and points to the cheap, lasting ones first.
The thinking behind it
BJ Fogg BJ Fogg, founder and director of the Stanford Behavior Design Lab, formerly the Persuasive Technology Lab, set the model out in A Behavior Model for Persuasive Design (2009), a paper at the Persuasive '09 conference. He wrote a behavior as the product of three factors, motivation, ability, and a trigger (the prompt, on this page), and held that all three must occur at the same moment or the behavior does not happen. The lab's first name, persuasive technology, says where the model comes from, and it reads as a designer's diagnosis: when the behavior is missing, find which of the three was absent at that moment, and fix that one.
How to run it
- Name the behavior exactly. One person, one action, one moment: log the expense the day it is spent, never "better expense discipline".
- Catch it failing once. Find a real moment when the behavior did not happen and ask which of the three was missing then.
- Time the task. Count the minutes, clicks, decisions, and things to remember; each one is a cost to ability.
- Shrink it until it works on a bad day. Cut steps and preset choices until a tired person on a busy afternoon would still do it.
- Put the prompt where the behavior happens. Tie it to a moment that already occurs, such as the card swipe or the end of a meeting, instead of a time on the calendar.
- Raise motivation last. Incentives and messages help only once the task is small and the prompt reliable, and even then they fade.
- Count the behavior itself. Track same-day entries, never survey answers about how much people care.
Where it goes wrong
Motivation goes first. A campaign, a leaderboard, or a bonus lifts the numbers for a few weeks, they slide back, and the team concludes that people do not care. Before any fix that depends on people wanting it more, ask whether a smaller task would do.
The prompt arrives at the wrong moment. A daily reminder at five in the afternoon reaches people far from the receipt. A prompt works when it lands while the person can act on it.
The task is shrunk past its purpose. An entry cut so far that it no longer carries what finance needs (a photo with no cost center) meets the count and fails the job. Check the output as well as the number.
When to use something else
- COM-B Model. When the gap may lie in the setting or in people's skills, and the diagnosis needs more than three questions.
- Implementation Intentions. When the person controls the behavior and needs an if-then plan more than a redesigned task.
- Forcing Function. When the behavior must happen every time, and the process can be built so the wrong step is impossible.
Where it sits in SOLVE eX
Go to the source
- A Behavior Model for Persuasive Design, Persuasive 2009, 2009