Let people experience core value before registration
Aliases: signup wall · gated registration · try before account
What it is
A registration wall collects identity before anyone has seen what the product does: an email prompt on launch, an account to read an article, a login screen after a few seconds of browsing. This entry is about timing. Core value should be experienced first; registration should follow a reason to stay. It is not about how many fields the form has, and it does not treat “let them try” as a durable guest identity with its own retention and merge rules.
Why it happens
People weigh expected value against the cost of disclosing identity. Registration is a bundle of hard-to-reverse disclosures—contact, password, sometimes a phone number—while the product is still an unknown good. The wall turns “see what this is” into “hand over an identity first,” so working memory holds a form and the value proposition never lands. People arriving with a task (a search hit, a shared link) are especially sensitive: the job was to consume or finish something small, and the wall rewrites it as opening an account. Drop-off there is not “too many fields”; it is “not yet worth filling.”
Studying it
Split the funnel into “reached the product” and “reached registration,” and compare walls placed before versus after value.
Independent variables: when the wall appears (on launch, at the first core action, after one successful experience), whether the blocked task is itself the core value. Dependent variables: depth of experience before the wall, leave rate at the wall, registration completion, next-day return among those who registered.
Lab participants told to “please register” almost never refuse the wall; live traffic or unincentivized remote tasks are required. A higher registration completion rate is not success by itself—if the denominator is only people who already decided to stay, the wall merely moved leave earlier. For content products, separate “finished reading” from “saved a piece”: the former is experience, the latter is the action that actually needs identity.
Where it stops holding
Services that legally require a real name before they can operate (payments, clinical intake, some communities) cannot push the wall all the way back, but they can still put scope, cost, and read-only explanation before the form. In enterprise SSO, “registration” is not an individual decision; the issue becomes whether the grant is explained. People returning from a paid receipt should be recognized, not walled again. Tools whose entire value is “your data is remembered” cannot demo without identity, so the wall may come earlier—if a template or sample is still visible first.
Applying it
- Name the core-value actions (finish a read, complete a calculation, generate once, see a result) and guarantee at least one can finish without an account.
- Bind registration to actions that must be remembered, synced, or notified—not to launch, the first screen, or arbitrary taps.
- When the wall appears, one line should point back at what just worked (“save the result you just made”), not a brand welcome unrelated to the task.
- Verify by walking a cold start to first value and noting whether registration interrupts; then measure how many leavers at the wall never completed a core action. That share falling is the timing actually moving later.