A/B testing | subscription flow | behavioral design
Paywall Redesign
Redesigning subscription paywalls and conducting multivariate testing to determine which design most improved click-through rates and subscription conversions.
Context and design problem
El Confidencial had accumulated multiple paywall variants over the years, many built by the subscriptions team rather than design. This led to inconsistencies across paywalls and with the broader design system.
Research Process
Audit of existing paywalls
A major issue with the previous paywalls was the large number of variants in use, combined with a lack of organization and oversight. To address this, we started with a comprehensive audit to identify which paywalls were currently active.
Benchmarking
We conducted an extensive benchmarking of paywalls used by digital newspapers from the United States, Europe, and the rest of the world. We divided these paywalls by modals, in-line, and banner paywalls.
Behavioral Design Principles
We explored key principles of behavioral design and carried out an exercise to identify how each principle could be effectively applied to the paywall design.
Key Questions
Modal, inline, or banner?
Each format had trade-offs in placement flexibility and how much information it could communicate to the user.
Which value proposition converts best?
Should we highlight features and benefits, showcase the subscriber experience, or make an emotional appeal like supporting independent journalism?
Direct payment or landing page?
Routing users to a landing page allowed us to show more detail, but added friction by increasing the number of steps to convert.
How much content to show first?
Should the paywall appear immediately, or after a portion of the article? If so, what's the optimal amount of visible content?
Personalized paywalls?
Should users from Google Discover with low subscription propensity see a different paywall than loyal, high-intent readers?
Microsubscriptions?
Could we offer single-article access via micropayment or as a registration reward, and what were the technical and financial implications?
Brainstorming and designing options
We designed options across desktop, tablet, mobile, and app to explore edge cases for each concept. Stakeholder reviews and user feedback helped narrow down the variants for our multivariate test.
Final designs selected for testing
Transactional Paywall
This version did was our simplest option, reducing complexity in the hopes that the user would go directly to the subscription landing page.
Simple Value Proposition Paywall
This paywall featured a straightforward value proposition, emphasizing the number of articles available under the paywall each week.
Social Proof Paywall
This paywall sought to nudge the user towards subscribing by using the behavioral design prinicple of social proof, indicating the number of other people reading the article.
Subscription benefits paywalll
This paywall emphasized the benefits of subscribing, including unlocking unlimited access, exclusive articles, newsletters, and the ability to comment on content.
Multivariate Test Results
The control paywall (simple price comparison between monthly and annual subscriptions) outperformed all tested variants, despite the research and strategic effort invested in the new designs. This unexpected result highlighted the importance of testing assumptions rather than relying on best practices alone.
