The idea
Compare a changed region with a credible comparison region, run a randomised price test, use a conjoint survey, or model transactions with controls for season, payday and competitors. Regional comparisons need evidence that the groups would otherwise have followed comparable trends. Surveys measure stated preferences; check them against behaviour. Randomisation helps isolate the price effect. Cover at least a monthly cycle and check whether stockpiling shifts purchases across weeks.
Use it thoughtfully
- Randomise who sees which price.
- Always keep a control group or region.
- Run tests for at least one full monthly cycle.
An example
ILLUSTRATIVE EXAMPLE: A test offers two prices to random groups through the app for five weeks, then compares revenue and margin per customer with a control group.
A question to try
Which of your last five price moves had a control group you could compare against?
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.