Back-End Metadata Pipeline
Optimizing Back-End data pipelines to power better customer facing experiences to increase conversions on iOS & AND.
Audible
UX Design
Data-Driven CX
Back-End Optimization
The Problem
The Opportunity
Product Detail Page tags aren't relevant enough and customers become frustrated going down an unintended discovery path.
Audible's back-end database functions linearly, when it comes to sub-titles or sub-genres (L2, L3) which can drive customers down to genres they didn’t originally intend or desire, disrupting a quality browsing experience.
Customers utilize the product detail page they are interested in to launch what we call “continued discovery” to find their next listen.
Enhancing the back-end meta-data structures and systems that power user discovery paths
The back-end optimization will build more relevant user discovery paths across major flows (Search, Browse, Content Discovery) as well as within independent components (recommended carousel, search filters, refinement)
Optimization Approach
Working closely with engineering it was discovered that we can applying specific service calls onto the meta-data elements. We broke them down into categories that help drive down the most relevant tags that are related to an Audio book title.
First Priority level: Relevancy
Second Priority level: Personalization
Third Priority Level: The combined "scores" of the first two
