Balancing registration and subscription in our Android first use experience
Role
UX Design, Visual Design
Collaborators
Product Manager, Android Engineers, Data Analyst, Copy Writer, iOS engineers (rollout)
Outcomes
72% increase in registration rates
100% increase in day 1 conversion
73% increase in 30 day halo conversion
$1MM increase in annual revenue
I was tasked with redesigning the soft regi wall on Android with the goal of improving registration rate without hurting subscriptions.
The existing experience asked users to choose between registering, subscribing, and skipping. My product partner wanted to keep this structure, while cleaning up the design up and making the actions clearer. I followed this approach, but also suggested an alternate solution.
This was the first project I completed as a product designer at The Times.
Pushing for a different solution
While exploring solutions, I found it was challenging to balance so many competing actions in one step. This is often an indication that we’re trying to give users too many options at once. If we don’t know which action we want the user to take, how can they know? I suggested that we separate the subscription action out onto its own step.
I worked with my product partner to design a test that would stay true to what she had outlined in her roadmap, while also testing the addition of a separate subscription step.
Working within technical constraints
On mobile and desktop web we usually show an exclusive offer right after users create a new account. This has been very effective at driving starts. However, in the apps we were unable to show dynamic pricing in the set up flow.
I designed a solution that focused on supporting journalism, and showed users some of our journalists in the field. We know from research and subscriber surveys that desire to support the journalism is a primary driver of subscriptions, when setting aside price.
I wrote an approximation of what I felt the copy should be and then worked with a copywriter to fine tune the language.
Variants
In addition to the subscription offer I designed two versions of the first step. These were designed to product’s spec of showcasing more account benefits.
Design of login and account creation screens were out of scope for this test
Results
Variant 1 (separate subscription step) won over control and Variant 2
Registration/login rate in both variants had a 72% lift over control
Variant 1 had a 100% increase in Day 1 conversion rate, compared with a 39% increase in Variant 2
Variant 1 had a 73% increase in 30-day halo conversion, compared with a 41% increase in Variant 2