A common intent shaping problem
Users skip every optional question
The short answer: let them try before you ask. Here is how to recognise this problem in your funnel, why it happens, and how the best consumer apps design it away.
How to recognise it
- Skip rates near total on optional flows
- Questions asked before the product has shown anything
- Answer quality as poor as answer rates
Why do users skip your questions?
A question's answer rate is mostly a function of when it is asked. Before value, questions are a toll: skip. After value, the same questions become tuning: answered, and answered carefully.
Users skip questionnaires from products they do not yet care about, which is every product in its first minute. The skip is not laziness; it is proportionate investment in an unproven relationship.
Forcing the issue backfires twice: mandatory questions get junk answers, and junk answers make personalisation worse than none.
How do you get questions answered? Demonstrate value first, then ask
Let the product make its case first: a sandbox, a sample, a first result on defaults. Let them feel what the app does at its generic best.
Then ask, framed as sharpening: “want this tuned to you?” lands entirely differently once there is a “this” worth tuning. Answer rates and answer honesty rise together.
Keep the ask proportional to demonstrated value: one question after one win, more after more. The relationship earns depth in instalments.
And honour the skip: defaults that work, no nagging, the questions available later in an obvious place. Some users tune on day five; the product should be ready when they are.
The building blocks that solve it
- Sandbox Experience: The product proves itself before it asks anything.
- Personalisation: Tuning offered as a payoff, not a toll.
Who does this well
Pinterest: The feed works on three taps: the deeper tuning happens after the product has proven the feed is worth tuning.
Spotify: No taste interview: the product plays first, learns while you listen, and asks nearly nothing.
Questions founders ask
Should any question be mandatory?
Only the ones without which the product literally cannot function, and even those deserve scrutiny: a default often works where a question stood.
Everything else converts better as an offer than an obligation.
What if the product needs data before it can show value?
Find the minimum viable question set, make each one visibly consequential, and show a taste of the output as early as possible, even a rough one.
A glimpse of the result funds the remaining questions.