Why People Don’t Buy Your Subscriptions - Issue 92
Appfigures Chat with industry leaders on metrics, benchmarks, and optimizing subscriptions
Hello analysts! Can you believe it’s May already? Are you prepared for the summer rush and seasonality to affect your metrics? I sure am. I just re-stocked my kitchen with Merlot.
One of the great things about being a blogger is that you are constantly being invited to fancy parties and fireside chats, all with the opportunity to connect with experts across many domains. The beauty of being an analyst-blogger is that you meet data leaders shifting the industry, and every chat turns into a sensitive and insightful conversation touching on experimentation, ROIs, KPIs, and other things that make analysts short of breath and quickened of pulse.
Today I am excited to share my recap of the most insightful discussion that caught my attention over the last few weeks - Appfigures Chat with Jake Mor, Founder & CEO of Superwall. Jake shared invaluable insights about the most important parts of optimizing your subscriptions and doubling conversion rates. He talks about what worked for his team to scale and grow FitnessAI, including tips on how to optimize the onboarding flow, benchmarks to follow, design concepts to use, experimentation, and more.
Today many apps and games can make a high profit on paid subscriptions without having a large business model. One of those examples is the FitnessAI app. Jake started the product from scratch and in just over 2 years has scaled it up to an empire that has revolutionized the fitness industry.
FitnessAI uses AI to generate personalized weightlifting plans. Based on 6M+ workouts, the AI automatically optimizes sets, reps, and weight every time you workout for optimal muscle growth and recovery.
Jake really understands the game of subscriptions, including how to turn a download into a paying subscriber (which is the goal for everyone).
How did FitnessAI start?
Jake is an iOS developer who created a simple fitness app back in high school. It was hardcoded with only his routine, and he would log his exercise routine every day. He then later created another app on top of it to let other people input their own routines and track their own workouts. In 3 years, he collected quite a lot of data that was enough to develop an algorithmic approach to weightlifting that would provide a user with a workout routine that they could follow, automatically adjusting sets, reps, and weights based on your and other people's workout history. He eventually got a team for FitnessAI and bootstrapped the app to a million in recurring revenue within a year. Easy.
Jake didn’t stop there. To grow FitnessAI, he had to constantly A/B test and iterate with onboarding flows and paywalls. He tried and used up many A/B test tools, and by necessity started another product and built an A/B testing platform for developers with respect to paywalls - SuperWall.
Getting people to pay you is more than just making great features
You may have great features, but people may not get to even encounter them because of a paywall.
When people land on your paywall, they might not know yet what they want (or what they’re missing!). So your paywall should be a marketing page. Don’t sell the features. Sell the experience.
App products fall into 2 groups:
One is when you offer the product for free, hoping people will subscribe later.
Another one has a paywall early in the funnel, and people subscribe without trying the product.
Both are challenging to grow, and Jake experimented with different paywall types and flows to end up with the most optimal combination.
Recommendations for how to optimize the paywall
Make everything remotely configurable
The first price and product offer iterations were not successful because only 8% of users were seeing a paywall. So, they had to invest in testing instrumentation and make the experimentation toolkit flexible and configurable. It’s not an option for every product, but if you address it early on, making a remote configuration will allow you to choose the paywall location, text, product, and audience, and continually iterate on it non-stop. After a few weeks of rapid experimentations, they got to at least 5% and 6% for Install-To-Trial. Over 1 year, they boosted Install-To-Trial up to 8%. After adding a video and experimenting more with the paywall location during onboarding, they got it up to 15%. That’s the impact of the ability and freedom to move the paywall anywhere.
Iterate with every step of the onboarding flow
Your ability to iterate is linearly correlated with how well your analytics is set up. You have to have the ability to read conversions in the onboarding funnel for every step. Create a funnel of user onboarding with a sequence of steps and learn their CVRs. Then remove steps with the highest drop-off rates.
The trick to optimizing the onboarding completion is to sort the onboarding steps by each step conversion rate. Put the highest conversion steps upfront, and the lowest converting steps at the end, and analyze them.
Be explicit, upfront, and transparent with your pricing
Don’t hide the paywall by “Continue” or other buttons. Instead, just be transparent about how much the subscription costs, what it includes, and the length of the trial. Being “shady” with your pricing might help you get higher conversions initially. But as the team noticed at SuperWall data, apps that offer transparent pricing have the same high trial conversions.
If people don’t convert on the paywall, you can offer them a survey asking why. For example, if people respond that they don’t want to sign up because they are afraid they won’t be able to cancel, you can simply add a graphic showing them how to do so, or be explicit that they can cancel any time.
Offering a survey asking why users canceled their subscriptions helps you with (1) understanding what doesn’t work and (2) retaining those users. You can offer them a promo or discount later on.
An experiment worth trying
Make your paywall the first screen people see. Users still can skip it and proceed with the onboarding funnel. If your bounce rate will be too high, make the skip button clear. With this screen, you will capture high intend users.
When users download your app, they’re doing so because they think it solves the problem that they have. They are likely to convert right away with a possibly higher price.
The sequence of optimization steps for the paywall:
Each gives about 20% up to 25% lift if fully optimized.
The location of the paywall is very important. The best practice is to show the paywall screen twice: the first time before the onboarding, and the second time after onboarding. The screen after may give you an additional 20% lift. Also, consider adding a paywall with every new app open event.
Regarding price tests, FitnessAI ran experiments with 10 SKUs in a month. Don’t be afraid to test prices. You might be surprised! Your CVRs might not decrease significantly.
Once a user has skipped a paywall, you can start offering them a discount. How do you know when it’s the right time to offer? Plot a histogram from Install to Conversion time with 30-minute buckets. Look at when 90% of your conversions happen. If it’s within the first day, it means that you should offer the discount after Day 1. If the majority of your conversions happen during the first 3 hours, then you’d offer a discount after the 3rd hour.
Industry benchmarks and conversions to aim for
Install-To-Trial rate is 20%.
Trial-To-Paid rate is about 40% to 50%. They have seen apps with 60%.
Paid user retention is 40% year-over-year.
It is easier to convert users who are using your app for free rather than new users.
If interested, watch the whole conversation here to listen to insights about attribution, LTV, Appfigures, analytics, and more.
Thanks for reading, everyone. Until next Wednesday!