Move the Paywall
Optimizing the payment step that caused 2/3 of users to drop from signing up for DigitalOcean.
Role: Lead Designer, strategy
Timeline: On & off Q1 2024
Team: Cross functional
What's going on during sign up?
Through multiple workshops I ran to evaluate the sign up and onboarding experience at DigitalOcean, the paywall step consistently came up as the biggest barrier.
The data showed that the payment set up page had a 64% drop off rate, higher than 4x the drop off rates of other steps. This means 2/3 of users make it through email verification, account creation and the onboarding questionnaire and abandon the process when asked for a payment method.
Why is this happening?
We can all relate to abandoning a sign up process when asked for credit card info. Through multiple channels, 3 main themes emerged as to why potential DigitalOcean customers were doing so.
“I want to explore first”
Evaluating DigitalOcean as an option so don’t have enough reason to hand over payment info before seeing the full offering.
“Why do you need payment info now?”
The info was there but consistent feedback asking why it was required during sign up.
“I tried but I couldn’t”
Even when users tried to enter one, they often encountered vague error messages or got locked out without a way to resolve it.
Let's move the paywall
Proposal: Let's move the paywall step later so users can explore DigitalOcean first.
My proposals included moving the paywall to resource creation or offering free credits, which DigitalOcean was doing on our website to new users anyways.
I also found that most competitors also require payment during sign-up but also free credits. We were asking for the same commitment as our competitors while offering less incentive.
Let's (not) move the paywall
Although stakeholders agreed 64% drop-off warranted a serious evaluation, there was strong pushback.
Revenue marketing
Risk to first month revenue metrics. Can’t justify budget to offer free credit for all new sign ups.
Security
Payment info is a critical security signal to deter bad actors and meet legal requirements.
Engineering
With an Ember to React migration happening, any solution in the console would double the eng effort and any new pages built in Ember was seen as a “waste”.
Previous experiments
- Users that sign up with free credits had higher churn rates.
- A prev. attempt at moving the paywall showed increase in drop off.
The strategy: 3 phases, 3 pain points
I understood the hesitation but I still strongly believed that addressing the drop off at paywall would bring the biggest value in the entire sign up process. To bridge this, I went back to everything I'd done so far: user feedback, workshop outputs and the constraints. I realized the pain points could be addressed in a phased experiment plan, ordered from lowest to highest in effort and risk but lowest to highest in value to users.
I presented this plan to my team and stakeholders and was able to get alignment with a key principle: we'd only advance to the next phase if the current one didn't move the needle. This would allow for exploring solutions while building a case for the bigger, riskier phases.
Phase 1: Better communication
Hypothesis
Making payment related info more prominent and clear will build trust and reduce hesitation.
Experiment
Improved copy and explored different visual hierarchies. Ran A/B test against a control for 2 weeks. (~40k users per variant)
Info scattered across multiple places across 3 pages
Info given contextually, with details in sidebar
0.01%
Increase in conversion
The problem wasn't what we were saying. Users weren't persuaded by better copy.
Roll out variant to 100% of customers and move onto next phase
Reflections
On pacing
I over-indexed on rigor in the early phases. Phase 1 could have been validated with lighter methods. Getting through earlier phases faster would have gotten us to Phase 3 sooner.
On phase 3
I would have tested a targeted a small SMB cohort with a $200 credit, tracking retention and first month revenue.