Learn how the Personalisation team created an infrastructure to support online experimentation and A/B testing for recommender models.
In this blog post, you’ll learn how the Personalisation team uses experimentation and A/B testing to improve recommender products. Find out why it’s important to measure the impact of recommender products on our marketplaces, what types of experiments they lead and how they created an infrastructure to support online experimentation at scale.
- Why experimentation is vital within any product lifecycle
- How the team compares activity and conversion metrics with and without new integrations
- Real-life examples of A/B test conducted in our marketplaces
- How the teams set up the experimentation platform using an API gateway and A/B service