A/B Test

Using data to find the top performing design.

Test Objective

Increase the number of new diners entering the funnel.


Incorporate ideas from the rest of the company. Design and test solutions to see which worked. Implement the winner.



A working session with designers and any others who wanted to participate came up with several dozen sketches. After the sketches were collected, the Team worked to filter them down to a manageable nine test versions.


Once the designs were determined, the everybody started on respective areas. Data would start planning the test setup. The designers would work directly in the code to style the elements after the engineers built the functionaity. QA would begin writing test plans.


After the test ran was completed, the team would review the results together and decide which version to implement. Ideally this is straightforward, but a "winning" test in this metric could negatively impact another critical metric.

Back to Case Studies