We moved the moment when a user is asked to input the credit card details from after the end of the free trial to the beginning of the free trial. This move reduced the number of signups to the free trial but significantly increased the free-trial-to-paying-users conversion rate, giving a significant uplift to the ARPU of the website.
Cooksmarts is a subscription-based business that gives its subscribers access to numerous meal plans, grocery lists and recipes. A user chooses a meal plan and Cooksmarts provides grocery lists and recipes for the selected meal plan. All you then have to do is buy the ingredients and cook. One thing less to think about.
Cooksmarts uses a free trial model where a user is offered to sign up for a free trial, and after the trial ends, the user is upgraded to a paid plan.
Our task was to increase the revenue per user (ARPU) metric by working with the conversion rate (CR) of visitors to paying users.
We started the analysis by evaluating the engagement rate of free trial users after signup. We confirmed our initial hypothesis, that users with a high engagement rate during their free trial have a significantly higher propensity to convert to a paid plan after the end of the free trial, as opposed to free trial users with a low engagement rate.
We know that one of the effective ways to increase an engagement rate is to persuade a user to pay or commit to a paid plan in the future. The majority of people use what they pay for. The whole subscription-based industry with regular payments is built around this notion. So if you want users to pay for your product, first make sure they regularly use it.
We reverse engineered this user behavior and decided to increase the CR to paid plans by stimulating user engagement during the free trial through a commitment to a paid plan in the future.
Our hypothesis: moving the moment when we ask a user for credit card details from the end of the free trial to the beginning of the free trial will stimulate engagement during the free trial and as a result increase the free-trial-to-paid-plan CR. We were, of course, conscious that this move would reduce the visitor-to-free-trial CR, but our focus is always on increasing the actual amount of money a business makes per user (ARPU).
Reducing the free trial CR by asking for credit card details to stimulate user engagement is a counterintuitive strategy, but it’s a calculated risk based on user behavior data.
We created an alternative flow, where we asked users to provide credit card details at the beginning of the free trial, as opposed to the control flow where credit card details are asked for at the end of the free trial.
To validate this hypothesis we conducted an experiment that measured the impact of the new UX on the visitor-to-paid-plan CR.
The experiment had the following characteristics:
Experiment type: A/B test
Traffic split: 50/50
Key metric: visitor-to-paid-plan CR
Number of sessions that took part in the experiment: 26,130 users
Number of conversions that were recorded during the experiment: 1,959 conversions
As was expected, the visitor-to-free-trial CR of the alternative variation dropped by 63%, but the free-trial-to-paid-plan CR increased by 91%. Overall this change brought an 86% uplift in visitor-to-paid-plan CR.
The engagement metrics of users of the alternative variation during a free trial, which indicate users’ propensity to convert to the paid plan, also increased:
The result showed a 95% statistical significance and power of 0.85.