Visitor-Type Blocks vs One Homepage Narrative A/B Test

Visitors with different needs who get lost on a single linear homepage leave marks in the data: first clicks scattered across unrelated pages in analytics, session recordings of long scrolls that end back at the top, and repeated support questions about team use. When the blocks make the paths clear, visitors pick the one that fits them and reach a relevant destination with less wandering. If you cannot narrow your visitor types down to a few clear options, settle that first; and if the destination pages themselves are the problem, a new homepage structure will not fix them.

Reach rate for relevant destinations

Compare visits to predefined relevant destinations per assigned visitor; where possible, determine relevance using independent information.

Also track

• Overall first-action rateMake sure overall progression does not fall.
• Return rate from a destinationTrack visitors who return from a destination; returning alone does not prove the choice was wrong.
• Bounce rateMake sure exits caused by uncertainty do not increase.
• Time to first actionReview time to first action alongside visits to target destinations.

• StructureReplace a linear introduction with fixed visitor-type cards.
• ScopeKeep destination content, offers and the rest of the page unchanged.
• OrderChoose the card order before launch and keep it fixed.
• Visual balanceUse one card layout; changing emphasis is a separate comparison.
• Traffic sourceReview routing results within acquisition sources.

1. 1Clicks on the visitor-type cards exist only in the variant, so they cannot decide the test. Compare the reach rate for relevant destinations across both arms instead; if the overall first-action rate falls or the bounce rate rises, a gain in relevant destinations alone does not justify shipping.
2. 2Findings here are bounded by this card order, this layout and the acquisition sources in the test, so read them within each source. If one card stood out through emphasis you did not plan, the result reflects that emphasis rather than the structure; test emphasis as its own comparison.
3. 3A return from a destination does not prove a wrong choice on its own, but check that visitors who enter the wrong block can easily switch to the right one. On mobile, make sure the cards have not shrunk until touch targets are cramped; if they have, the result measures tapping difficulty, not the structure.
4. 4If the two structures end up close, some visitors may be helped by the clear paths while others are lost to the extra decision. Review time to first action alongside destination visits. Check with the A/B test significance calculator whether the gap in reach rate for relevant destinations is beyond chance; if it is not significant, do not call the new structure a winner.