When it comes to mobile app marketing, the importance of customer lifetime value (LTV) cannot be underestimated. Imagine you could reach users that are more likely to perform meaningful actions within your app such as completing registrations, tutorials or, ideally – making in-app purchases.

The advantages of forecasting customer lifetime value go beyond just being able to build your customer acquisition budget and funnel accurately. It lets you sharpen your marketing strategy and adjust it according to a channel, medium, GEO, creative, et cetera.

Let the Numbers Talk

mobile app industry statistics

It’s no news that the mobile app industry is fierce: there are more than 6.5 million apps out there, and most of them are never even discovered.

Whether your app succeeds or fails is in the hands of your end users. Research shows, 90% of mobile apps fail within the first three months. The plot thickens when you hear the following statistics:

  • 77% of daily active app users are lost in the three days after the install
  • There’s a drastic decline in numbers when you look further down the funnel: actual in-app purchases stand at only about 2%

Marketers must understand that not all customers are created equal to ensure app longevity and success.

customer lifetime value graph

The Time to Focus on Quality is Now

More and more industry players understand that to achieve scalable and sustainable growth and success they need to measure everything to be able to forecast yield per user.

The current economy – which has urged businesses to become more customer-centric – together with the latest developments in the mobile industry, are pushing digital marketers and app developers to put more value on quality rather than quantity.

Look at Facebook and Google, for example – the two companies that have already recognized the importance of customer lifetime value. Facebook now allows you to create value-based lookalike audiences, pinpointing users with the highest value for you (those that are likely to download your game app, for instance) to maximize your campaigns worth.

As for Google, it has recently rolled out Universal App Campaigns (UACs), which are all about using machine learning algorithms to locate customers that have the highest possibility of downloading and interacting with your mobile app.

The secret lies in the correct understanding of user behavior and capability to predict it.

Customer Lifetime Value in Action

Customer Lifetime Value Process

Let’s face it: there’s a short window of time for catching the user and understanding their intent: what stage are they at in the funnel – is it someone who is looking for information or someone who wants to make a purchase? As a marketer, you’d want to make sure that user sees the right ad at the right time.

  • Audience segmentation – when predicting LTV, it’s best to, first and foremost, look at your audience segments: everything from user demographics to the type of browser they use. When done correctly, user segmentation lets you create buyer personas relevant to your business as well as for each step of the marketing funnel.
  • Modeling – The next step would be to turn to quantitative prediction models, which analyze your audience segments to determine the kind of creative/offer that your audiences would find most engaging.

Machine learning and data mining algorithms empower the most common prediction models, which need to be tweaked and sometimes customized to ensure they match the business situation at hand.

There are also probabilistic approaches to modeling, the most common of which is the Pareto/NBD model. The model works on the principle of ‘a coin and a dice’: it uses purchase history to understand whether a customer is to churn (coin) and then tries to find a pattern using the Pareto distribution. The dice principle is used to examine the purchase rate and determine the number of orders a customer can make. Then, the Pascal distribution is used to model the dice.

When checked over time, both machine learning and probabilistic approaches tend to generate outcomes of similar quality.

Here’s a detailed example our CEO, Inbal Lavi, gave at The Marker’s Digital Conference:


  • Automation – this stage entails integrating prediction models and user targeting methodologies into a system that would use A/B testing to run your campaign most efficiently: knowing exactly when to launch or stop it or deducing what kind of relevant creatives/offers should be served to users to reach maximum ROI.

Bottom Line

Today, smart digital marketing comes down to knowing when, where, why (users’ intent) and what kind of content you need to target high paying customers and ensure your app durability.

Here at Webpals® Group, we exclusively operate thousands of top-ranking sites and use our SEO, SEM, social, mobile, native advertising know-how and proprietary technology to deliver high-value end users to global businesses and promote their mobile apps. We do all that in various verticals like e-Commerce, food delivery, finance, gaming, live streaming, and more.

We can track and analyze the behavior of different audiences anywhere. Using statistical models and AI-based technology, we pinpoint homogeneous audiences with specific intent and target their lookalikes on a diversity of digital platforms.

It’s the uniqueness and quality of the seed audience that drives a much better eCPA to our partners and allows us to successfully operate on a performance-based marketing business model even when it comes to mobile app promotion.

Contact us, and we’ll make your mobile app the fittest to survive in the ever-changing, hungry-for-innovation marketing world.