Buy box at the bottom of the listicle


Hypothesis
By adding a buy box at the bottom of the listicle page, we expect visitors who have consumed the full listicle content to convert without navigating to a separate product page, thereby reducing funnel drop-off and increasing overall conversion rate.
What we learned
What we tested: Whether adding a buy box at the bottom of the listicle page, so visitors could purchase right there instead of clicking through to a separate product page, would lift conversion versus the original listicle that routes visitors to the product page. What happened: The original page (control) won. The buy-box version performed worse on the metrics that matter most. - Conversion rate: Control 3.2% vs Variation 2.9% (down 10.0%) - Revenue per visitor: Control $4.13 vs Variation $3.84 (down 7.0%) - Average order value: Control $129.43 vs Variation $133.62 (up 3.2%) Statistical confidence: There's about a 92% chance the original page is the better version on conversion rate, and roughly a 75% chance it's better on revenue per visitor. The slightly higher average order value on the buy-box version was not meaningful. Why it matters: The hypothesis was that letting primed visitors buy at the bottom of the listicle would cut funnel drop-off and lift conversion. The data points the other way, keeping the click-through to the dedicated product page converted better, so embedding the buy box in the listicle hurt rather than helped in this run. Note: This was the re-launched run and lasted only ~5 days (Jun 4–9), below the usual 7-day window, so treat the magnitude with some caution. It's also a redirect-based test (control stays on /7-reasons-why; the variation redirects to /7-reasons-why-offer and a -bogo-v2 page), if those destination pages differ from the originals beyond the buy-box placement, that could partly affect the read.
Want an agency that shows you its losses?
Every agency has losing tests. We're the ones who publish them. Free teardown: we walk your store live and show you the three highest-impact tests we would run first.