The idea
DellaVigna and Linos (2022), cited in the supplied reference, compare average take-up effects of 8.7 percentage points in published academic studies with 1.4 points in government nudge-unit trials. That gap is evidence from those settings, not a universal one-sixth conversion factor for marketing forecasts. Use the comparison to motivate realistic expectations and a controlled local test.
Use it thoughtfully
- Set effect expectations for your own setting.
- A/B test every nudge against a control.
- Measure uncertainty before promising a rollout outcome.
An example
ILLUSTRATIVE EXAMPLE: A team expects a new default to lift renewals by 2 points, not the 10 points a book suggests, and tests it on 10% of customers first.
A question to try
What effect size did you assume in your last behavioural campaign, and where did it come from?
A useful framework is a starting point, not a guarantee. Effects vary with the audience, product and situation. Use truthful evidence, respect customer choice, and test the result.