With Predictive Insights on Likely Buyers, 8fit Improves Conversions and Customer Engagement

8fit, the health and fitness app offering customized workouts and meal plans, wanted to improve conversions and customer retention. 8fit worked with Braze to predict which app users were most likely to convert, then targeted this group with special offers, leading to 3.75X higher conversions.

While today’s workout and wellness programs are increasingly accessible everywhere, thanks to mobile devices, it’s still important for brands in the space to be smart about how they communicate with and encourage their users. For 8fit, which offers workouts, customized meal plans, and meditations via its app, customer engagement is critical to its success: The app is available in six languages, and has 40 million downloads and counting.

To maximize the impact of its CRM campaigns, 8fit wanted to increase their number of paid subscribers without offering big discounts to all of their users—but they weren't sure who to target in their marketing campaigns to reach that goal. By leveraging the Braze platform to send targeted subscription offers to users based on their likelihood to purchase, 8fit was able to achieve higher conversion rates while sending fewer messages.

Predicting Customers Most Likely to Convert

8fit has consistently worked to optimize its customer retention and conversion efforts. As part of that effort, they found that one key to reducing customer churn was to encourage new customers to become app users: Some 8fit users subscribe on the web, so it is important to have them download the app, proceed with onboarding, and engage with the product in order to persuade them to stick around. 8fit chose to work with Braze because of the platform’s ease of use, and because marketers have all of their tools for reaching customers—via email, push, and in-app messages—in one place.

8fit worked with the Braze product team to use Predictive Purchases to better understand how likely their users were to make a purchase. This feature, part of the Braze platform’s Predictive Suite, uses machine learning (ML) to find relevant patterns in the behaviors of users who have made a purchase in the past, and uses those insights to predict which users will make a purchase in the future. Users are assigned a Purchase Likelihood Score from 0 (least likely) to 100 (most likely), which 8fit leveraged to target relevant audiences with personalized paid subscription offers. By offering lower subscription discounts to users with medium to high Purchase Likelihood Scores—and saving the big discounts for less likely subscribers—8fit successfully improved overall campaign ROI.

Using email, push, and in-app messages, 8fit tested subscription offers at various discount levels in targeted messages to 1.6 million users. The discount amount was determined by whether a user had a low, medium, or high Purchase Likelihood Scores. Over a series of several campaigns, 8fit tested different messages and channels with different offers, such as emails linked to web subscriptions compared to emails that opened up in the app’s paywall. 8fit also tested message efficiency by excluding users who were least likely to make a purchase from receiving these messages.

"Working with the Braze platform and team is a continuous learning curve for us," said Stefan Clausing, VP of Marketing at 8fit. "This is especially true when participating in an Early Access program that allows us to be at the very forefront of product innovation. Purchase Prediction not only helped us to understand our customers better but also enabled us to instantly act on these insights without further engineering efforts.This way we were able to increase the efficiency and effectiveness of our CRM campaigns."

8fit Results: Increase in Conversions

With Braze and Predictive Purchases, 8fit can query audiences based on users’ expected future behavior and see how these users respond to different channels and offers. When 8fit marketers create CRM campaigns, they can easily target users with promotions that are specifically tailored to their purchase intent. In addition, the campaigns give 8fit insights into the user behaviors and attributes that are correlated with their conversion likelihood, helping marketers create future campaigns for greater success.

In one A/B test, 8fit saw that conversions for users with high probability to purchase were 3.75X higher than a randomly selected cohort of the same size. In another multivariate test, 8fit increased the relevancy and efficiency of their campaigns by excluding low-scoring users, which saved sending 100,000 emails weekly with no negative impact on conversions.

“The new Purchase Prediction feature enables us to add impactful strategies for our overall growth," said Marlie Vermeeren CRM Manager at 8fit. "Working closely together with the Braze Product Management and the Customer Success team gave us more knowledge and insights about our customers and their conversion behavior, as well as how to utilize it most effectively. Now that we better understand how to optimize conversions, we are confident that we’ll be able to retain even more 8fit users over the long term.”

Final Thoughts

Giving away a “free lunch” by sending the same offers and messages to each and every customer and prospect is something all brands want to avoid. Understanding better how to target messages and promotions requires insights about which customers are most likely to respond positively to offers. For brands intent on growth and customer retention, knowing the difference between likely and not-so-likely conversions is critical to improving ROI.

To learn more about choosing messages, content, and timing that drive conversions, read the Braze article, “Go From Browsing to Conversions With In-App Messages.”