Design Sample Data for New Accounts

Onboarding & Growth · onboarding, empty states, feature discovery

Updated 2026-07-28

Give new accounts something to look at. An empty product cannot demonstrate itself, and most people cannot judge a tool from a blank canvas.
Make the example realistic. Sample content should look like a real project, because placeholder nonsense teaches nothing about what the product is for.
Mark it unmistakably as an example, so nobody mistakes it for their own work or reports it as a bug.
Make it removable in one action, and say that removing it is safe. Users will not commit real work to a workspace they think is cluttered.
Let people explore it safely. Sample data is the ideal place to try deleting, sharing, and editing without consequence, so do not lock it down.
Prefer importing the user's own material where possible. Their own data is more persuasive than any example you can write.
Match the sample to what they told you during signup. A generic example after a set of qualifying questions wastes the answers they gave.
Retire it once real content exists, and never let it distort onboarding metrics or usage reports.

An empty state explains what to do, and sample data shows what it looks like when it is done, which is a much stronger argument.