About the client
The client is an Australian online retailer, selling candles and gifts. They receive approximately 15,000 to 17,000 visitors per month, mainly via email, Google ads and Facebook.
The average amount of products purchased in one order was quite low. The store wanted to increase the average order value without decrease overall conversions.
Conversion Optimisation research
Looking at the store’s analytics, most of their sales came from one particular brand of candles and diffusers. With a couple of fragrances within that range performing very well with users often ordering both a candle and a diffuser of the same fragrance.
A/B Test Details
The test was run on a range of the top selling products that had a diffuser / candle combo available in the same fragrance.
The test was run on both desktop and mobile devices.
Test Results
Revenue results
- The test collected 6155 sessions over 25 days.
- Version A (Control). Sessions 3125. Revenue per session $7.99
- Version B (Variant). Sessions 3030. Revenue per session $10.37
- Version B won the test with 99% likelihood to be the best and revenue per session uplift of 27.7%
Transactions were used as a guardrail metric. We wanted to increase average order value and revenue, but not at the expense of overall conversions.
To my surprise the conversion rate didn’t go down, but actually increased by a substantial amount.
Transaction results
- The test collected 6155 sessions over 25 days.
- Version A (Control). Sessions 3125. Transactions 211. Conversion rate 6.75%
- Version B (Variant). Sessions 3030. Transactions 237. Conversion rate 7.82%
- Version B won the test with 94% likelihood to be the best and conversion rate uplift of 15.8%
This result had the potential to increase revenue for the client by about $180k per year.