Novelty effect

The novelty effect is a short-term boost in usage or engagement that appears simply because something is new and draws attention, not because it delivers lasting value. Users click, explore, or interact more with a changed interface, a new feature, or a redesign for a period after launch, and that lift fades as the change becomes familiar.

It matters most in experiments measured over a short window: a test that runs for only a few days may report a treatment effect that looks strong purely because it is novel rather than better, and the same lift is unlikely to persist once the change stops being unusual. The mirror image is "change aversion," where existing users react negatively to any change regardless of its merit, which can just as easily understate a genuinely good change in its first days.

Practitioners guard against novelty effects by running experiments long enough to see the metric stabilize, by tracking the trend of the average treatment effect over time rather than a single endpoint, and by comparing new users, who have no baseline to be surprised by, against returning users, who do. A feature whose lift decays steadily toward zero over the test window is a signature worth flagging before a permanent rollout.