TripAdvisor's iPhone application was a series of repetitive and tedious prompts to sign-in, sign-up, turn on location services, and turn on notifications. And if you skip some, there's another system prompt that asks you to reconsider the decision. I was tasked with rethinking and redesigning the onboarding experience, and making sure that the flow was frictionless.
I collaborated with one of our product managers early on in the quarter. In our initial kick-off meeting, we documented as many entry points users come through and all the possible flows we could take them through. My concern was conflicting business goals vs. user goals. I felt very strongly that based on my competitive analysis and research on onboarding best practices, that we should aim for a more organic, contextual onboarding experience.
If you're interested in exploring some of these features more in-depth, please download the TripAdvisor® iPhone® and Android® applications.
Because I wanted to ensure that the flow was frictionless, we decided to test a number of different areas in the app to get permissions, next destination for a trip, and a user's home location. Some tests were run during in the initial app open (full screen interstitials), some were done directly in the homescreen, others more contextually on list views or after second app opens.
I did user testing and questionnaires with participants to document their propensity to turn on permissions, what concerns they have with them, how frequently they travel or use a product like TripAdvisor, and whether or not they understood the value propositions presented in the prototypes I built.
Below you will see the process and flows for some of the onboarding tests we ran. After doing about a dozen different tests, I ended up with two different versions I wanted to test with users.
This onboarding flow was designed to give users a clear value proposition when presenting them with notification and location permissions. We wanted to preface the iPhone system prompts with a better explanation as to why we need these permissions.
Here is a prototype I built in Principle. It helped me validate decisions regarding the copy and visuals for the test. Given that this is a common experience with mobile applications, we made the decision to quantitatively test it and evaluate the results.
We saw a ~6% lift on location opt-ins and a ~1.5% lift on notification opt-ins.