Skip to content

Technique · Influence

Self-prediction prompt

Ask users to predict a near-term action they already want to take ("Planning to finish the first lesson today?" with Yes and Not today), then make that action easy at the moment they named. Saying yes to a prediction is a small commitment that raises follow-through.

Evidence grade B. Moderate evidence: solid studies with some limits.StartReturn

Plain label: ask users to predict their next stepAlso known as: question-behavior effect, self-prophecy, RSVP prompt, will you finish today?

How it works

When people are asked whether they will do something desirable, most say yes, and the stated prediction then pulls their behavior toward it so they stay consistent. The effect is small but reliable, and strongest for actions people already value. Turning the prediction into a guilt lever later ("you said you would") breaks the sense of free choice that makes it work.

In the book. When people are asked to forecast whether they will do something worthwhile, more of them then do it (Ch.3, pp.51-52). INF-T-CC-03 (Influence Ch.3, p.51)

In a product [S]. A one-tap prediction right after a first win, tied to a reminder the user schedules.

Evidence Evidence grade B. Moderate evidence: solid studies with some limits.. Reliable but small in meta-analyses.

Evidence warning

What the research says

  • In the original study, residents asked to predict whether they would volunteer for a charity later volunteered at a far higher rate than residents simply asked to volunteer (a reported rise of about 700%).

    Influence Ch.3, p.51 (Sherman), 1980

  • A meta-analytic synthesis of many studies found that asking people about a future behavior changes it, with a small average effect.

    Spangenberg, Kareklas, Devezer & Sprott, Journal of Consumer Psychology, 2016

Before you start

  • The value event is tracked.

How to build it

Mobile
After the first lesson, ask "Will you do Lesson 2 tomorrow?" with Yes and Not now; on yes, offer a reminder at a time the user picks.
Web
At the end of onboarding, ask how many sessions the user plans this week and show progress against their own number.
Email and SMS
For events and webinars, ask registrants to confirm they will attend and offer a calendar add on the same screen.

What can go wrong

  • Quoting the prediction back in cancellation or guilt messages.
  • Asking for predictions about actions users do not care about.

Pairs well with

  • Pairs with F21 if-then plans: the prediction is stronger with a time and a place.
  • Pairs with in-t-cc-09 for longer goals.

Limits

What we don’t claim

Sources

  • Influence · Robert Cialdini

    Ch.3, p.51

  • On the self-erasing nature of errors of prediction · Journal of Personality and Social Psychology, 1980

  • A meta-analytic synthesis of the question-behavior effect · Journal of Consumer Psychology, 2016

    Evidence grade B. Moderate evidence: solid studies with some limits.

Source IDs: INF-T-CC-03

Book content is paraphrased, never quoted at length. Last verified October 24, 2026.

Your product

Find the stage your product is stuck in

Gamify It reads your product, picks the one move that matters at its stage, and sizes the test to your traffic. Connecting is free; a Pro subscription starts it.