Mobile game maker keeps big spenders engaged

A mobile game maker wanted to increase player engagement. By using Pecan, they were able to cut the player churn rate in half with models built in a matter of days.

person playing game on mobile game



prospected VIPs


lower churn rate


uplift of spend 
per user

The challenge

Keeping and enticing gamers

A mobile game maker wanted to start using predictive analytics for its in-game management and to assess its business performance and bottom-line impact.

The company, like any online company, is interested in identifying its customers, building distinct profiles and matching each of them with appropriate offers and content. Prior to using Pecan, the company only managed to partially leverage its data to improve business results despite boasting many unique active players a week and with hundreds of thousands of weekly transactions. Like other online companies, it also suffered from player churn. Other challenges included difficulty to adapt the offering to different player profiles and inability to identify high-value players early enough in the process.


A predictive approach

Before turning to Pecan, this gaming company used other statistical and manual methods to build distinct profiles and match each of them with appropriate marketing offers, content and products.

The results were less than optimal so the company built with Pecan 3 deep learning-based predictive analytics models:

  • To identify potential high rollers. 
  • To predict which of these customers is susceptive to marketing activities. 
  • To predict which of these customers will churn – and how to prevent that.

This type of data-driven segmentation is valuable to the business, as it can be used to ensure that players are continuously getting value and as such increase revenues, reduce churn and help to build successful future products.

This company used the Pecan platform to build the models in a total of 12 days, compared with 4-6 months required by alternative solutions, which also consume huge and expensive resources of data scientists.

Using Pecan instead of traditional methods meant that the cost of the project was estimated at 25% of alternative solutions.

Furthermore, the AI models were built by one of the customer’s BI managers using the Pecan platform, avoiding the need for trained and expensive data scientists.


Owning the future

Compared to the standard identification of high rollers (or VIPs), the model identified a population of more than 200% additional potential high rollers. Treating them as such early in their lifecycle and with Pecan’s recommended treatments meant that the company managed to convert the majority of them into highly valuable customers. Each segment was tested separately and the average segment saw 3X uplift of spend per user.

With users that were predicted to churn similar tests were made and the results were as strong: 

  • Users that weren’t treated actually churned
  • Pecan’s recommended treatments help cut churn in half
  • Treated users spent more than 4X more than those who weren’t treated

This mobile gaming company determined the results of this campaign: less churn with higher spent per user and, most critically, low to no effort – the company empowered its BI team with the capability of building state of the art AI predictive analytics models, without having to hire data scientists for the job.

Want to get started with predictive analytics?