Fietspunt Improves Its Product Page To Increase Purchases

FietsPunt set out on the path of A/B testing with the aim of increasing the number of online orders.

The original homepage looked impressive, clearly displaying key information such as shipping and returns policy right in the top banner. But FietsPunt’s product pages showed a high bounce rate that naturally affected sales adversely.

This is how control looked:

FietsPunt’s owner Roeland Van Oostenbrugge had read another VWO success story in which the client saw a 58% increase in sales by adding a Trust Pilot widget. As he put it, “The inspiration for starting this test came from some VWO cases. It was funny to see another VWO case that was almost similar to this”.

Roeland felt that visitors were apprehensive about trusting a small and relatively unknown company with online shopping. Having read about the six principles of persuasion posited in Robert Cialdini’s classic book ‘Influence’, Roeland decided to conduct an A/B test with a variation that included a Facebook widget on the product pages that showed more than 1,000 “Likes”. This, he believed, would constitute “social proof” for visitors to the website.

The test was set up and run using the VWO platform. Half the traffic was directed to the original page, and the other half to the variation. This is how the variation with the Facebook widget looked:

However, the test didn’t seem to have much impact on user behavior, because the problem of bounce persisted. He tested another variation where the widget was placed at a different location- but that did not change the situation either.

An undeterred Roeland, who believed that his hypothesis of providing social proof was right (but the way he was doing so was not), selected another tool from the credibility arsenal- a TrustPilot widget that showed the latest customers reviews. This test too was set up and run on the VWO platform. He tracked “add to carts” and “sales”. The A/B test aimed to test if live customer testimonials helped allay visitors’ anxiety and boosted trust (and hence sales).

The widget was added on the bottom right corner of the page. This is what the variation looked like:

The test was run for 15 days on around 16,000 visitors before Roeland decided to implement the changes on the website.

The version with the Trust Pilot recorded a 36.73% increase in orders and had a 99% chance to beat the original.

When asked how he expects the increase in orders to impact revenue, Roeland said, “I assume the turnover to grow linearly with the conversion rate. In that case, our monthly extra revenue would be approximately 90,000 Euros ($122,660).”

It is interesting to note that the same underlying persuasion principle of ‘social proof’ delivered such widely-varying results in the two tests. The Facebook widget showing “likes” didn’t make any difference to conversions whereas the TrustPilot widget showing customer testimonials did have a major impact. It is unfair to directly compare the two tests as they were run several months apart. Nonetheless, it is useful to understand why the variation with customer reviews worked better.

We believe there was a difference because the TrustPilot widget provided greater reassurance to visitors and prospects. Some 63% of consumers say they are more likely to purchase from a site if it has product ratings and reviews. Around 70% of online shoppers in US say they look at product reviews before making a purchase. The TrustPilot widget displayed reviews by customers who had already made the purchase. This assured prospects that other people were actually buying the products, thereby removing the last barrier for conversion to take place.