We redesigned recommended products block on PDP.
Result: 7% growth in revenue per visitor (RPV).
Our client (www.bigl.ua) is an eCommerce marketplace with 11+ million monthly visits (according to Similarweb.com). The monetization model is pay-per- view of PDP. Sellers pay for each pageview of their product pages (PDPs) on www.bigl.ua.
The goal of the CRO project was to increase the overall website revenue per session by increasing average number of PDPs visited per session.
We started the CRO research with calculating the impact of each stage of the funnel on the overall website revenue and its growth potential.
The most common user journey involved users landing on a category page, then clicking on a product in the listing to be redirected to a product detail page (PDP).
In order to prioritize our optimization efforts, we calculated a share of each UI element that leads to paid clicks throughout the funnel in overall website revenue.
We found that a significant amount of revenue (6.1%) was generated by clicks on recommended products block at the bottom of the product detail pages – “Similar products from the same seller” and “Similar products from other sellers” blocks.
According to heat map analysis, only 7% of PDP visitors scroll to the point where they see these blocks.
Moreover, a high number of clicks on the “Show more” call-to-action (CTA) below the recommended products indicated that users are keen to see additional product recommendations.
Our first hypothesis was to move the recommended products block higher on the page so that more users will see it and click on it.
The only available place that we found on PDP where we could move the recommended products blocks was occupied by sticky block with “Buy” CTA that was always visible on the right-hand side of the page, starting from the fold…
… and till the recommended products block.
According to the data from previous experiments the sticky block with “Buy” CTA significantly increases conversion to transactions. So we couldn’t completely replace this sticky block with recommended products block.
So we decided to move the sticky block with “Buy” CTA to the top of the screen replacing the sticky search bar.
Redesigning PDP so that recommended products block is positioned vertically in the middle of the page will increase the visibility of the recommended products block and lead to higher CTR of the block.
To test this hypothesis we moved the recommended product block to real estate occupied by sticky block with “Buy” CTA and sticky block with “Buy” CTA we moved to the top of the page.
Sticky header with “Buy” CTA instead of search
Similar products from the same seller block higher on the page, instead of sticky “Buy” CTA
Variant A (control) – the “similar products from the same seller” and “similar products from other sellers” blocks are placed at the bottom of the page.
Variant B – “similar products from the same seller” recommended products block moved upwards on PDP.
Variant C – “similar products from other sellers” recommended products block moved upwards on PDP.
The experiment ran on all desktop traffic in A/A/B/C format.
Variatoin B increased CTR of “similar products of the same seller” recommended products block by 342.99% and “similar products from other sellers” recommended products block by 10%. Overall website revenue per session increased by 5.95% and conversion to purchase by 7.86%
298,071 visitors took part in the experiment over 14 days, and the results achieved 99% confidence level based on not sampled data.
Performance of A–A test ensured the data quality. The client’s CRO team repeated the experiment on live site and had the same results. The variation B was implemented as the new control.
– Priorities your hypothesis based on the impact of change on site’s revenue.
– The impact of some top-performing elements could be multiplied simply by increasing its reach and visibility. For example, moving elements that genarate revenue from real estate where only 15% of audience see it to real estate where 100% of visitors see it will predictively increase your revenue. Simple math.
– But always define the secondary metrics that should not drop and A/B test its change to understand long-term impact on revenue. For example, your experiment may be successful in increasing CTR, but may cause an increase in bounce rate. In that case, you can plan next alternative version to eliminate this problem.
Schedule a call with our analyst.
He will learn about your business and provide you with actionable recommendations