A quiz app sounds trivial: ten questions, a score, a share button. The build is trivial. The distribution is not.
The problem with quizzes is that one player is not a user
A quiz has an awkward shape as a product: the person who makes it gets value from the results, not from taking it. That means the creator is not the audience — their friends are. Every design decision follows from that.
What worked:
- The quiz has to be sendable as a link, not as an app install. Anything that requires the recipient to sign up dies at the first hop.
- The result page must be legible without context. If a friend opens a result and cannot tell what it was about, the chain stops.
- The creator's name goes in the title, not the description. People open things addressed to them.
Answering is the funnel, not the landing page
My first version had a polished landing page and a slow quiz. Analytics showed almost everyone bounced on the landing page. The version that worked dropped the landing page entirely: one button, straight into question one.
Every question is a micro-commitment. Ten questions is roughly the limit before drop-off becomes visible in the data.
Scoring is where people argue — use that
Scores that are too generous feel fake, and scores that are too harsh feel personal. The version people shared most was the one where the result told them something specific about which answers they got wrong, rather than just a number.
What I would skip next time
- Accounts (not needed until people make multiple quizzes)
- A gallery of public quizzes (empty for the first month)
- Custom themes (nobody opened the editor)
If you are building a friendship quiz, design the share path before the question set. The quiz is the easy part.
Notes from shipping and then shrinking the same app twice.
Top comments (0)