What was your initial criteria for selecting a replacement solution and did this criteria change as you delved more deeply into the process?
We’ve worked with Apptus since 2004, but, as is often the case in technology, we weren’t on the latest version of their software. We talk regularly so we knew that their behavioural search was a big leap forward. It was clear to us that the new search solution had to include a behavioural feedback loop. With a catalogue of 12 million products we needed to rely on behavioural data in order to present the most relevant product for each specific search of our customers.
What sparked the need to change your search, merchandising and personalisation solution? Was it simply time for a redesign or were your existing technologies no longer fit for purpose?
CDON.COM has significantly increased the number of products it sells, and the number of categories had exploded. Given that 30% of our visitors use site search, with such an increase in the size of our catalogue it was clear that our existing search was going to need an overhaul.
The risk was that we’d have gone to the trouble of extending our range only to discover customers were no longer able to find what they were looking for. The current search was also customised for our legacy product categories such as music, movies, books and games.
When you sell these and combine it with clothes, shoes, sport and white goods it requires a more intelligent search to present the most relevant results.
What was the most important criterion which made you re-select Apptus?
Of course, we do keep up-to-date with the market, but Apptus were the only supplier to include a behavioural feedback loop in their search solution that could cope with our traffic load and product catalogue.
However, we like to be methodical and decided to test Apptus’ latest search solution against what we were then using. We always test all our new features on the website since we don’t know how our customers will react and behave.
Sometimes we think a new feature is a clear improvement but when tested on our customers it has proven otherwise. We understood the theoretical benefits of the behavioural feedback loop in delivering better, more relevant search results but we wanted to put an actual figure to it and so we conducted an A/B test.
50% of visitors were offered the old search solution, and 50% the new.