SaaS Onboarding A/B Test Examples: 8 Results

Of 7 measured SaaS onboarding tests, 6 won and 1 lost.

The most useful pattern: show new users exactly what to do next and get them to a real result fast, with checklists, progress bars and their own data on day one.

The wins come from AURUM, VWO, DashThis and Brainshark case studies, and they mostly moved activation and trial-to-paid conversion.

The one loss, from OnSwitch, shows that holding value back to force users to return can cost sales.

Seven tests is a small set, so use these as starting points for your own onboarding tests.

What a SaaS onboarding has to do

Onboarding sits between sign-up and the first moment of value.

Its job is to turn a new trial user into an active one, and an active one into a paying customer.

That is why most tests here measured activation, setup steps completed, or free-trial-to-paid conversion, rather than sign-ups.

Most-tested changes on the onboarding

• Checklists: a visible list of setup steps (AURUM, VWO).
• First-run redesigns: a guided first experience with a progress bar (AURUM).
• Clearer setup steps: fixing the step where users drop off (DashThis).
• In-app messaging: tips, upgrade offers and a setup wizard (Brainshark).
• Access rules: locking content to drive return visits (OnSwitch).

Biggest measured wins

AURUM ran three tests on the Astrea trial.

A floating onboarding checklist lifted activation rate +187%, with key setup events up +21% and +25%.

A first-run redesign with a progress bar lifted activation +105% (+7 percentage points).

Showing real data on day one of the trial raised engagement by +16 percentage points (+125% relative).

A VWO case study found the same thing with a "Get Started" checklist for free trials: +127% free-trial-to-paid conversions.

That is the strongest sign here that onboarding changes reach revenue, not just engagement.

DashThis made the integration step in onboarding clearer and saw +50% onboarding funnel completion and +25% first-step completion.

Connecting data sources is often where trials stall, so a clearer integration step is worth testing for any product that needs one.

Brainshark's in-app tips, upgrade offers and setup wizard brought +150% enterprise free trial signups.

Tests that lost (and why)

OnSwitch tried locking free content for three days to push users to come back.

Users did return slightly more, but they bought less.

The likely reason is that holding back value delays the moment users see what the product does, which is the opposite of what the winning tests did.

The lesson for SaaS teams: measure onboarding tests on purchases or paid conversion, not just return visits or engagement.

A metric that rises can still hide a drop in revenue.

An onboarding checklist built from the results

• Give users a visible checklist of the few steps that lead to value.
• Show progress so users know how close they are to done.
• Use real data early: fill the product with the user's own data on day one where you can.
• Find and fix the stall step: often an integration or connection step.
• Guide in the app with tips and a setup wizard rather than relying on email alone.
• Don't lock away value to force return visits.
• Measure the right outcome: activation and trial-to-paid, not only sign-ups or logins.

Tools to run these tests

Onboarding tests usually run inside the product, so you need a tool that can split users by account and track in-app events such as setup steps and activation.

Feature-flag and product experimentation tools suit this better than page-level testing tools.

Whatever you use, define activation before the test starts and give it enough trial cohorts to reach a clear result, since trial volumes are often lower than website traffic.