Outcome Testimonial vs General Praise A/B Test Idea

Survey answers that call the testimonials too generic, or support contacts asking what problem the product solves, suggest testimonials do not give visitors enough to decide. Visitors who read an outcome close to their own situation should then head for the primary action with more conviction and not drop off at the next step. Confirm that every specific outcome is in the customer's own words and that you have permission to show names, titles or organization details; if visitors leave before they reach the testimonials, look at that point first.

Primary action completion rate

Compare whether the type of testimonial changes completion of the target action.

Also track

• Post-action drop-off rateMake sure visitors drawn in by a specific testimonial do not drop off at the next step.
• Testimonial interaction rateTrack measurable interactions; visibility does not prove the testimonial was read.
• Testimonial visibilityMeasure whether the section enters the viewport.
• Return or cancellation rateTrack later losses that may follow a mismatch between claim and experience.

• SelectionCompare genuine general praise with genuine testimonials that describe a specific outcome.
• PresentationKeep testimonial count, placement, design and identity format unchanged.
• IdentityWhere possible, choose comparable customer contexts; the speaker's identity can also affect credibility.
• Separate comparisonsTest testimonial count or identity detail later.
• RelevanceReview results by the use case and industry represented.

1. 1If post-action drop-off or the return or cancellation rate rises, a gain in the primary action completion rate gives you no case for shipping. If a specific testimonial sets expectations the experience does not meet, the losses show up later, and a testimonial that presents an unmeasured figure as exact makes that risk larger.
2. 2This pair of testimonials and the use case they represent set the limits of the finding. If the testimonials come from a single industry, do not generalize the result to other industries, and since a different speaker can also change credibility, do not credit the gap to the outcome framing alone.
3. 3Before searching industry or use-case groups for a hidden effect, check whether each has enough data. Work out how many visitors each group needs with the sample size calculator; differences in small groups are easy to misread.