Predictive Churn Prevention

Churned players affect us all. At Fast Track, we are lucky to have our very own data model to assess and predict player behaviour.

Players can leave your site at any time without any obligation to tell you. Luckily, the Probability of Return is a predictive model to keep you one step ahead. As a user of FT CRM, you have this intelligent predictive data model at your disposal. Use it to bring some powerful intelligence into your churn prevention strategy.
Schedule automated campaigns with an incentivisation strategy. Contact your players when they start showing signs of inactivity. Repeat that with another offer once they become less and less likely to return. Losing players will become a thing of the past.
  1. ​What Benefits Will You See​
  2. ​How Does the Model Work​
  3. ​How Do We Build Predictive Segmentation​

🏆 What Benefits Will You See

Many of us have CRM Activities scheduled for players that have not engaged with our brand for a while. Say, for example, if they have not deposited in the last five or ten days.

Personalisation

But not every client is the same. Some customers play after their payday a couple of times a month, whereas others play several times every week. With these differences, it's harder to know when a player has churned, and when do you need to contact them?

Redefine Churn Prevention

Luckily, predictive models can help. The Probability of Return model redefines how we look at churn strategies. It analyses players' past depositing behaviour to predict their future behaviour. The model calculates the frequency of players' Activity and gives them a score every 24 hours. This score will determine how likely a player is to return, or, how likely the player will churn.
React swiftly to prevent churn while avoiding unnecessary spam.

📩 How Does the Model Work

  1. A player has shown a pattern of depositing every Tuesday and Friday.
  2. Suddenly the player stays inactive for one full week.
  3. The data model recognises this change of behaviour.
  4. Due to the change, the model assigns the player to a new Segment.
  5. This change is the catalyst qualifying the player for automated churn prevention campaigns that you should build.
  6. Players will receive a personally targeted campaign encouraging them to return to your site.
Below, you can find another simple example of how it could work for a highly active player.
Example of Automated Churn Prevention
Example of Automated Churn Prevention

A how-to guide on how to get started with your Activities:

https://resources.ft-crm.com/tutorials/predictive-churn-prevention/building-automated-churn-prevention

💭 How Do We Build Predictive Segmentation

The model calculates a score for each player that runs from 0 -100%. According to their score, each player is automatically added into a Segment that can change every day.
In total, we have added seven Segments into FT CRM. The first group consists of players that are the most likely to stay active (Sky-High). From there, they go down to groups of players that are less and less likely to remain active all the way to churned players (Dead).
Predictive Segments and Scores
  1. Sky-High - Score is above 83.5%
  2. High - Score is between 67% and 83.5%
  3. Medium - Score is between 50.5% and 67%
  4. Low - Score is between 34% and 50.5%
  5. Frail - Score is between 17.5% and 34%
  6. Critical - Score is between 10% and 17.5%
  7. Dead - Score is less than 10% (Inactive)
To find which players belong to each Segment, open the Active Players Dashboard under the Predictive Player Insights. You can find it under the 'Insights and Analytics' tab on the left-hand navigation menu on your instance of FT CRM.

More information about the Dashboard:

Active Players Dashboard
/knowledge-base/insights-and-analytics/dashboards/system-dashboards/active-players-dashboard