How the redesign of a product page increased its conversion rate by 164%
Enzo is the name of a brand-new wardrobe avaliable for sale in our E-commerce. It is a good material product with a pretty look, but somehow its sales wasn´t going very well.
Field research: meeting Enzo
I went to a store where Enzo was in the showroom for customers. We took some photos and learned more about the product benefits with the salesman.
In this research We were able to confirm that Enzo was indeed a good quality product. It has sliding doors, a inner smart division to keep your clothes organized, a big mirror on the door and a strong MDF structure. However, these features that make the product so special, were not being highlighted enought on the product page in Toque a Campainha’s e-commerce.
So our hypothesis was that the users wasn´t seeing Enzo’s features because of the product description layout. To support it, we made a checklist and verified other points that we consider important for a product to be salable on the internet:
- Images – high quality and detailed;
- Price – good for a closet, above the average price of the category;
- Delivery time – fine, the product was in stock;
- Technical specifications: width, height, depth, and weight specified;
- Product description design – not so good.
Sketch and Designs
After the checklist, we decided that it was time to make something concrete with these information, so I started making some sketches and then a layout proposal for the product page:
New layout proposal
The following layout was added after the fold, below the description area.
A/B Test: validating the hypothesis
With the new layout ready, it was time to validate the hypothesis. The A/B test was designed to verify the effectiveness of a richer visual designed for the product description content. The results would tell me if the new version of Enzo product page would increase the conversion rate and consenquently the revenue.
The test was set only to desktop users and the method used was to split the traffic by 50% to each version.
To support the experiment, a heat map of the 2 versions was also generated.
As we expected, the variation would generate more interaction from the users because that the page height was bigger than the original. But the interesting fact was that the common area between those 2 versions, product description, that had also a higher interaction.
After 16 days of experiment, with a 98% of trust rate, 3.000 sessions and 23 transactions, the hypothesis was validated.
The original version sold 6 wardrobes against 17 of the variation.
The new version proved to be 163,9% greater than the original.