Meeters
Checkout (how we crashed it)
Team: 1 PM, 1 Designer (me), 2 Developers
Responsibilities: research, prototype, design, test, quality assurance
Time: 3 weeks (design), 2 months (development)
Situation
In Meeters, users can purchase experiences as virtual tickets and a digital subscription plan. The checkout flow was made of a total of 5 steps.
Cart page
Login/signup page
Participant details
Recap
Payment
We found that there was a significant drop in the login page.
The current login and signup screens included an email input that looks for existing accounts in the DB. If the email is new, it goes to the signup page.
What I did
First I sent a survey to our users to find out their familiarity with external authentication providers.
Out of the 94% of participants that already used social login in the past:
50% use it sometimes
38% use it often
12% use it always
Goal
The product decision was to introduce a signup option using Google and Facebook. This way we wanted to reduce friction during the signup flow and prevent the accidental creation of unwanted accounts.
The participants in the user test were chosen among our users that already had an account. During the test:
74% of participants used an external authenticator to signup
26% signed up with email and password
Some errors were generated by a CTA that was standing out too much and they were solved before sending it to development.
Thoughts
My assumption here was that if the majority of people use social login sometimes, over time it might have been the preferred login method. This was based on the fact that I designed many flows to synch existing accounts and because providers like Facebook simplify the registration process by prefilling input fields.
I didn’t consider that the complications that we were bringing to the screen might have offset the benefits (Spoiler: this is what happened)
Then I designed a prototype that I tested with real users.
Results
3 months after the release of the new login/signup page we found out that the conversion rate of the tickets had dropped significantly after the introduction of the new signup screen 😱.
Not just that, we also saw that more than 50% of new users were still using email and password to create an account.
In the next case study, I’ll show you how we fixed it ✌️
What I learned
Looking at this case study now, I believe that there have been a few reasons why the first implementation was unsuccessful:
Wrong priorities: the decision to develop social login options was not backed up by a real user need. Simplifying the registration of a new account might have had a bigger impact on the user journey. Instead, we focused on a feature that we wanted because we normally use it.
We should have spent more time understanding if our users would actually use it, considering that adding more choices comes with an increase in users’ decisional fatigue.