30 November 2016

Learn tips from Memrise to increase in-app conversions with pricing experiments

Posted by Tamzin Taylor, Partner Development Manager at Google Play, & Kristina Narusk, Head of Production at Memrise

Getting people to install your app is one thing, getting them to sign up to your paid offering is quite another. It's important to understand the complete journey your users take from installing your app to paying for something. Once you do, you can experiment on the flow to try and increase conversions. Memrise has found great success in experimenting on their language learning app to increase the number of paying users.

Four experiments Memrise use to improve conversions

Memrise makes languages fun with a number of different learning modes you can play to help increase your vocabulary in a chosen language. You can download the app for free and play some of the modes or take advantage of their premium subscription offering called 'Memrise Pro' which offers new game modes and additional features like offline learning. Memrise recently ran a number of conversion experiments with the main objective of increasing the Average Revenue Per Daily Active User (ARPDAU). These experiments tested multiple user experience and pricing experiment scenarios.

1. A/B test how messaging different user benefits can impact conversion

What they did: Memrise wanted to know what motivation and call to action would convert the most users to buy a Pro subscription from a locked game mode in the app. To do this, they ran an A/B test with two similar designs, featuring different reasons for the user to upgrade, and compared the results to their original upgrade messaging.



Screen Shot 2016-10-11 at 15.26.00.png
Screen Shot 2016-10-11 at 15.25.34.png
Test A: Focus on ‘difficult’ words with an orange background.
Test B: Focus on ‘favorite’ words with a pink background.

Results: Test A performed the best. Conversion to Pro in Test A was 28% higher than in Test B. Pro mode usage was subsequently 9.7% higher in Test A compared to Test B too.

Next steps: After seeing how test A won the experiment, Memrise applied this creative across the board. Subscribers driven by that particular mode increased as a percentage of all subscriptions in the app by 16%. Memrise plans to run additional A/B tests at others points of conversion in the app to see if they can increase the results even further. They'll also try different text for the call to actions.

2. Test whether adapting to local price points results in sustainable uplift

In 2015, Google Play launched new minimum local price levels in countries around the world. To take advantage of the new price points, Memrise tested lowering localised prices in certain markets to better match purchasing power. Prices were an average of 6 times lower during this experiment.

Results: After 30 days, Memrise saw the following changes in conversions to paid users:



🇹🇷
Turkey
180%
🇧🇷
Brazil
182%
🇷🇺
Russia
99%
🇲🇽
Mexico
115%
🇮🇳
India
5.1%
🇮🇩
Indonesia
152%
🇰🇷
South Korea
120%
🇹🇭
Thailand
70%
🇲🇾
Malaysia
27%

Next steps: The change in price affected the subscription dynamics with more users taking advantage of Memrise's in-app discounted offer in most countries. The offer was for annual subscribers only and has led to a positive effect on LTV. One insight from the experiment was that Indian users prefered to have the option to subscribe in weekly or monthly increments and not just annually. Memrise is still tracking carefully to see whether the discounted subscription pricing will lead to an increase in conversions.

3. Test when and how often you offer free trials to see if that affects conversion rate

Memrise occasionally offers users, who aren't Pro subscribers, a free trial of one of the Pro game modes while cycling through the various free modes. After the free trial session, users are presented with an offer to subscribe. Memrise experimented with the offer's timing making it appear more frequently while users were cycling through normal free sessions Instead of after every 49th session, users saw the unlocked mode after every 21st session.



Screenshot_unlocked.png
Screenshot_unclocked_offer.png
An example of a free trial of a Pro mode.
After completing a free trial, users see a discounted subscription offer.

Results: Offering a free trial more frequently paid off. The conversion rate increased by 50% while all other conversion rates remained the same.

Next steps: Memrise maintained the more frequent offer cadence and has seen revenue growth as a result.

4. Test whether seasonal discounts result in more conversions Memrise launched a 'Back to School' campaign presenting all users with a discounted annual plan offer for a week in September 2016. The aim was to convert more users and generate higher value users from annual subscription plans.

Screen Shot 2016-10-11 at 16.03.14.png

Results: Memrise saw two effects from the seasonal offer. As a result of only presenting an annual period and removing weekly and monthly, 20% fewer users per day converted to Pro. However, because more people were taking an annual subscription than a shorter subscription, the average revenue per day increased by 32% justifying the change.

Next steps: Memrise plans to test different offers in the future with a combination of subscription offerings. They'll also focus offers in countries like Turkey and Mexico, where they saw the biggest increase in conversions.

Keep experimenting and take advantage of new features to improve the user experience and increase conversions

At Playtime San Francisco, we announced that introductory pricing for subscriptions would be coming soon and the feature is now live. By continually testing messaging, pricing, offers, and free trials or discounted trials, you could increase the conversions in your app and your ongoing revenue just like Memrise. Learn more about Google Play in-app billing subscriptions and get the Playbook for Developers app to stay up-to-date with features and best practices that will help you grow a successful business on Google Play.

How useful did you find this blogpost?