Open almost any product critique and the conversation gravitates to onboarding — the first screen, the empty state, the tour. It makes sense as a starting point, because first impressions are legible and easy to test. But a disproportionate amount of design effort ends up parked there, on the five minutes a user spends once, while the screens they'll actually live in for months get whatever attention is left over.
Retention isn't decided on day one nearly as often as it's decided on day two, or day seven — the moment someone has to choose, unprompted, to open the thing again. That decision is shaped by what they remember being useful last time, not by how polished the welcome screen was. A beautiful onboarding flow that leads into a confusing or empty second session is a beautifully wrapped box with nothing memorable inside it.
The habit-formation literature — Nir Eyal's 'Hooked' is the most cited version of it — frames this as a loop: a trigger gets someone to open the product, an action gives them something, and a variable reward makes the loop worth repeating. Most teams design the trigger and the first action carefully and leave the reward, and the reason to come back a second time, mostly implicit.
In practice this shows up as underinvested surface area: empty states that just say 'nothing here yet' instead of prompting a next action, notification copy written by whoever was free that sprint, a returning user's home screen that looks identical whether they've used the product once or fifty times. None of it is deliberately neglected. It's just further from the part everyone reviews in a design crit.
Our practical version of this: sketch the day-two screen before the day-one screen. If you can't picture what a returning user sees and why it's worth opening again, the onboarding flow you're polishing is optimizing for a session that, on its own, doesn't predict anything about whether the product survives.