Guarantee Bullets vs Paragraph Copy A/B Test Example

If visitors keep contacting support about topics the page already answers, heatmaps show the paragraph being skipped, or session recordings catch them moving past the section quickly, the assurances are probably buried too deep in the paragraph. With bullets, they should find the assurances more easily, ask fewer questions on those topics and reach the primary action with less hesitation. Do not set the list up as a stand-in for the legal text, and if the problem lies in the assurances themselves or in missing terms, review the content first, because a format change will not fix it.

Primary action completion rate

Compare completion to see whether a bulleted reassurance list changes the target action.

Also track

• Related support contactsCheck whether questions about the listed topics decrease.
• Section visibilityMeasure whether the section is seen.
• Time on pageRead time alongside completion without assuming extra reading is helpful.
• Return or cancellation rateTrack later issues caused by misunderstood assurances.

• FormatTurn the existing reassurance paragraph into a short bulleted list.
• Keep constantKeep claims, conditions, order, placement and other content unchanged.
• UnderstandingAssess relevant questions or research separately from conversion.
• AudienceReview new and returning visitors.
• Separate testsNew reassurance claims, a different bullet count or adding icons are separate tests.

1. 1Hold any gain in completed primary actions against the return or cancellation rate. If the list brings more actions but more returns or cancellations later, assume an assurance was misunderstood and do not treat the gain as a reason to ship, and do not treat a rise in time on page as helpful reading by itself.
2. 2Keep this result scoped to this page and this set of assurances. Review new and returning visitors separately, and check that the list does not come before information needed for a decision, such as price or coverage, and that section visibility stays similar in both versions.
3. 3Do not close an unclear test on the first weeks of data; returns and cancellations show up later, so keep tracking them through the follow-up window. Enter the current primary action completion rate and the lift you want to detect into the sample size calculator, then divide the sample per variant by daily traffic per arm to see how many days the test needs.