Designing for the second visit

We over-design first runs. There is a whole genre of case study about the welcome flow, the progress dots, the confetti at the end. Almost none about the moment three days later when someone opens the product again and has to remember why they liked it.

The second session has no momentum

On the first visit the user arrives with intent. They clicked an ad, read a recommendation, decided to try the thing. That intent carries them through a surprising amount of friction. By the second visit the intent has gone and been replaced by something much weaker: mild curiosity, or an obligation they set themselves.

So the second session has to supply its own reason. Usually that means showing the user their own work — the note they started, the project half-configured, the data that arrived while they were away — rather than the product's features.

Empty states are a second-visit problem

A first-run empty state can be aspirational, because the user has not invested anything yet. A second-visit empty state is a report card, and if it still says nothing here yet after a real attempt, the product has told the user their effort produced nothing.

The fix is rarely more encouragement. It is finding something true to show: a partial result, a related example, a next action sized for the five minutes the person actually has.

A cheap diagnostic

Look at your analytics for the ratio of second sessions to first sessions, then look at how much design time went into each. In most teams I have worked with the ratio of attention is roughly inverted from the ratio of importance. Moving even a fraction of the onboarding budget to day three is the highest-leverage reallocation available to most products.