Buyakilt.com Lifts Conversions by 31.6% and Increase Revenue by Six Figures Using Convert Experiments
CLIENT BACKGROUND & CHALLENGE
Macdonald Sporrans runs several eCommerce websites including Buyakilt.com, a unique online kilt retailer in the UK. Fergus Macdonald, Founder and CEO of Macdonald Sporrans, wanted to increase conversions on his product page and hypothesized that testing security badges would lead to the desired outcome. The greater success in this case was not the simple test idea, but rather the ease of setting up, implementing and measuring the results.
All what was needed to begin the test was Convert Experiments a simple £20/month subscription.
The A/B Badge Tests
There are several locations on an eCommerce website where a security seal would be helpful, the most common being the cart and checkout pages.
vs. vs. none
Buyakilt.com already had a Comodo badge in the sidebar so for their first test, they added a Norton by VeriSign badge in the top right of the shopping cart, leaving the existing badge in the bottom left sidebar. This did not improve the conversion rate.
The second test was the checkout page, where they added a Norton by VeriSign badge on the top right. That test resulted in a 17% lift in orders placed with a confidence of 99.29% over a period of five weeks. So Norton by VeriSign conclusively won over having no logo on the checkout page.
Because optimization is an ongoing process, Buyakilt.com didn’t stop after their first win. They immediately dove into a third test to determine if there was a difference between using the Comodo and Norton by VeriSign logos on the checkout page in the top right corner. After four weeks, the Comodo badge won with a 12.49% lift with 89.06% confidence over the Norton by VeriSign badge. Even though the confidence level was below what is typically recommended by our tool, they were thrilled with the results and confident enough that they hardcoded on the checkout page.
The series of two winning test gave a total of (1.17×1.1249) 31.6% lift in orders placed over a period of just nine weeks. Even though the results were not all with the recommended confidence level, they impacted the business by a whopping six figure increases in revenue.
Original: Checkout Page Without Badge
Variation 1: Checkout Page with Norton Badge
Variation 2: Checkout Page with Comodo Badge
From a technical perspective, this was quite a simple implementation. All Buyakilt.com did was install the Convert Experiments code on their site and add the badge to the main template. They then used the visual editor to select the specific badge and “remove” the entire badge again.
They ran it on 100% of the visitors going to the checkout and subsequently, on the card and measuring the goals (sales) using the URL matching goals.
A/B Testing Results
Here is a breakdown of the tests and their results:
Not only has this experiment proven that it’s worth test badging, but it’s important to test different ones to have a better understanding of trust perceptions. Thrilled with these results, Buyakilt.com implemented the same changes to their other websites http://buyahipflask.com and http://buyatankard.com.
- 12 Nov, 2012
- Posted by Dennis van der Heijden
- 2 Comments