Audible
A web experience on Amazon.com
Project Details
Overview: Successfully initiated and led the design of a new onboarding flow for Audible audiobooks on Amazon.com by introducing an experience that makes choosing a first audiobook less intimidating. With this flow, downloaded content during the trial period increased from 76.5% to 92.1%, which in turn increased Audible’s conversion of trial members by 10.6%.
The Problem: On Amazon.com, 30.3% of users don’t listen to an Audiobook during their free trial period. A lack of interaction during the free trial has a direct impact on whether a user will convert to a membership.
The Problem: On Amazon.com, 30.3% of users don’t listen to an Audiobook during their free trial period. A lack of interaction during the free trial has a direct impact on whether a user will convert to a membership.
Research & Defining the Problem
Amazon has an extensive backend system of metrics tracking, and Audible has a clear funnel of customer feedback. While I was looking for a problem to solve, I came across data that only about 70% of users that sign up for a free Audible trial listen to an audiobook with it. Digging deeper and interviewing users, I found that a common reason for this is an "analysis paralysis". With so many options available, and being new to Audible, users didn't want to "waste" their credit and therefore never chose a book.
Solution
Working with the data science team, I found that they had recently developed a new algorithm. This algorithm searches for keywords within the actual written text of a book, and determines what the book is actually about (rather than surface level searching of titles). For example, searching for dragon would make connections to books about castles & knights.
Using this tech I designed an onboarding system where users could enter keywords or book titles they have read before, and we could recommend titles that they would actually be interested in. You can view a mobile prototype of this same flow here.
Using this tech I designed an onboarding system where users could enter keywords or book titles they have read before, and we could recommend titles that they would actually be interested in. You can view a mobile prototype of this same flow here.