Experiment walkthrough
Testing a sticky CTA on mobile.
I worked on a mobile-web experiment that kept the primary action visible while visitors scrolled. This walkthrough explains the idea and a general approach to evaluating this kind of change.
Company details and underlying experiment data are omitted. Visuals are illustrative. The measurement section explains a general evaluation framework.
Make the next step easier to reach.
A visitor may decide to continue while reading further down a page. If the primary CTA is outside the viewport, taking that next step can require scrolling back to find it.
The hypothesis
Keeping the primary action visible could make it easier for visitors to continue when they’re ready.
Two ways to present the primary action.
Illustrative comparison. These are not screenshots or reproductions of the original experiment.
Results.
-
+13%
In conversion rate
-
≈ $2.9M
Estimated annual impact
Measure the completed conversion.
The intended conversion should guide the decision. CTA clicks help explain behavior, but they do not establish whether the whole journey improved.
-
Completed conversions
Use completed conversions per eligible visitor as the primary outcome.
-
CTA clicks
Track use of the primary action to understand whether visibility changes engagement.
-
What happens next
Check whether additional clicks lead to progress through the funnel or more drop-off.
Check the tradeoffs.
A persistent CTA takes up screen space. Check that it leaves room for the information and controls people need to make a decision.
- Does it cover important content or other controls?
- Does it work on smaller screens and when text is enlarged?
- Can people use it with a keyboard and assistive technology?
When this is worth testing.
Consider this pattern when visitors need to read or compare information before taking an action, and the primary CTA can move out of view.
Review the page and funnel evidence before deciding whether visibility is the right problem to address. Unclear value, pricing confusion or friction later in the journey may deserve higher priority.
Have a conversion problem worth investigating?
I help turn conversion problems into experiments worth running, working with your existing team.