Inside AppMetrica
March 6 2024

Maximize your mobile marketing ROAS with LTV & Churn Predictions

New features in AppMetrica: LTV & Churn Predictions

Gone are the days of waiting several months to assess the Return on Ad Spend (ROAS) and optimize mobile marketing campaigns weeks after they’ve been launched.

AppMetrica’s new LTV and Churn Predictions feature empowers mobile app user acquisition and product managers with fast insights about app users, their lifetime value and churn probability from Day 1 of launching your campaign and getting the first app installations.

With LTV and Churn Predictions, you will be able to:

  • See what ad channels you can boost or turn off on the Day 1 of your ad campaign
  • Attract a high value audience to your mobile app based on accurate LTV predictions
  • Optimize your best-performing ad campaigns to grow your revenue and ROI
  • Maximize your ROAS and understand what channels to invest in the most
  • Analyze and compare cohorts of users based on their LTV and churn probability
  • Identify a segment of users with a high likelihood of churn and retain them

Learn more about LTV and Churn Predictions in AppMetrica

LTV Predictions

Lifetime Value (LTV) is a crucial metric in mobile app marketing. It represents the total revenue a user is expected to generate throughout their entire relationship with the app.

AppMetrica’s LTV predictions take this concept to the next level by leveraging AI to find potential users with the highest LTV. This feature allows user acquisition managers to optimize their strategies by targeting users who are likely to yield the highest returns.

What’s under the hood

LTV predictions evaluate each user within 24 hours of them joining the app and form an LTV prediction for 28 days that follow. Based on these forecasts, you can send postbacks into your ad network directly from AppMetrica and optimize campaigns for topLTV users with just a few clicks.

In contrast to classic optimization suggestions based on traditional metrics like ‘time spent’ and ‘engagement’, the new AI-based predictive model collects and analyzes massive amounts of data around every user’s potential LTV to find the highest quality leads for your ad campaigns. To ensure accuracy, it uses the formula

LTV = Revenue All x P (LTV > 0) + Revenue (first day) x 1 — P (LTV > 0)

LTV predictions also let you segment users into various LTV cohorts (like top 5, top 20, top 50 and bottom 50) and compare them.

To test the accuracy of the predictions and evaluate which method attracted a more engaged audience, the founding team ran an A/B test for a gaming app:

The test compared two options:

  • Optimization based on the «User played for at least 10 minutes» event
  • Optimization based on the predictive model for the top 20% of paying users (top20LTV)

The campaign settings and budgets were the same. The first ten days of the campaigns were spent learning and building up the data set. Within the next week, AppMetrica experts collected installation data to use in the predictive model.

AppMetrica’s A/B tests showed an overwhelming advantage of optimizing campaigns based on the LTV reports. Read the full story on how Yandex Games used LTV & Churn Predictions to increase their user engagement by 10,5% in this case study.

Churn Predictions

User churn is a common challenge in mobile app marketing. Identifying users who are likely to leave the app and implementing proactive measures to retain them is vital for long-term success.

How it works:

AppMetrica’s Churn predictions enable app owners and marketing teams to identify new users who are most likely to churn over time as soon as they install the app.

The AI model analyzes all active users over the period of 3 weeks, scoring their activity on a daily basis. While the model requires no specific metrics, it accurately predicts users who are more likely to quit using your app depending on the typical user’s lifecycle in the particular app. The generated report splits all users in groups based on churn probability: >95%, 75-95%, 50-75%, and <50% with a 99% accuracy rate as guided by the 3-sigma rule.

By leveraging predictive analytics, this feature empowers you to implement targeted retention strategies to prevent user drop offs — such as launching personalized push notifications, personalized incentives and more.

Knowing which users are more likely to churn gives you an upper hand at engaging them and preventing their leave from the app as soon as possible. AppMetrica lets you reach your users via personalized push notifications directly from your account.

Read our step-by-step instructions on how to use LTV and Churn Predictions in this How-to Guide and watch video instructions.

Try LTV and Churn Predictions for your mobile app campaigns

Reach out to our customer success managers to get a free trial or upgrade your subscription today.