New one way flight segementation - A/B test

Hypothesis: Offering the customer a dedicated one way cross sell would increase conversion because we don't know which leg of their journey they are on and we want to cater for all possible scenarios.

The problem: From analysing the data it was noticed that one way flight customers were converting poorly compared to return flights. User testing revealed that the location we were showing was not always the required one which could cause confusion and not obvious on how to edit the search.

Iteration 1: Having defined the 'problem' to be fixed, concepts were produced and agreed internally. The concept grouped the cross sell into 3 top level filters (outbound location, inbound location and a search all locations). From here an A/B test was setup to test this with actual users.

Results: After 4 weeks of testing we got the confidence level in the results. Interactions increased 15.8% and car hire conversion increased 31.9%, with these results we pushed out the test 100%.

Iteration 2: From analysing the data it was noted that 55% of the bookings were coming through the inbound leg. From this another iteration was put forward using tabs and defaulting to the inbound (sketch attached), this would bring cars back to the cross sell rather than filters for the leg of the journey.

A/B test on going (iteration 1 versus iteration 2)...

Further iterations: Having the inbound, outbound and new search tabs to follow through the flow to the list view of all cars, giving the customer an easy way to change the search for their one way trip. When a new search is completed this would be cached in another tab (saving their last search).