Building Automated Churn Prevention
In this section, we go over instructions on how to include the predictive model's insights in your daily campaigns.
🌟 How Does Predictive Segmentation Work
Fast Track has already included ready-to-use segmentation attributes into your Core Segmentation Model. That means that you will find these attributes in your regular list of segmentation criteria. We will go over detailed instructions on how to use them later in this Tutorial.
For now, two main principles are relevant to understand. First, is that you can select which predictive segment they belong to today. Second, you can see if there was a change; in other words, if the players belonged to a different segment yesterday than the one they belong to today.
Primary Considerations for Segmentation
- Which predictive segment player belongs to today?
- Was there a change between today's and yesterday's segment?
- Was the change positive (player is more likely to return)?
- Was the change negative (player is less likely to return)?
🥇Take Full Advantage from Automation
With the predictive segmentation attributes, you can automatically target the players that are becoming less likely to return to your site.
Let's say you have a player who, based on the model's calculations, today has a Low probability of coming back. But yesterday they were Medium probability. You can schedule a bonus campaign for these players to fire every day at 17:00, each time they fall back to the Low segment.
For full automation benefits, run your campaign unlimited times per player. That way, you can always reactivate each player the moment they start falling off the wagon.
Keep your players engaged longer.
🚀 How To Set Up Segments
Next, we will go through the segmentation values you will need. We will also present an example of a set-up that combines all the relevant fields.
Which Segmentation Values to Use
Segment Field | Change in the State of Engagement |
Description | Change in the State of Engagement will reflect the numeric change from the previous segment. For example, -1 would mean the player has gone down one step in the segment chain and is less likely to return to your site. |
Variables | -6 to 6, from negative changes to no change at all (0) to positive changes |
Segment Field | State of Engagement Today |
Description | Current Probability of Returning Segment - this represents the likelihood a player will return (or not) to your site today. |
Variables | Dead, Critical, Low, Frail, Medium, High, Sky-High |
Segment Field | State of Engagement Yesterday |
Description | Previous Probability of Returning Segment represents the likelihood a player was going to return (or not) to your site yesterday. |
Variables | Dead, Critical, Low, Frail, Medium, High, Sky-High |
Depending on the set-up on your instance of FT CRM, the segment names might be different.
You might see, for example, PAlive Current Segment, and Probability of Returning (Change).
They will work in the same way as we have described above.
Configure the Segment
We have added a simple example set-up for one churn prevention segment below. It will target a High-Value player that falls to segment Low. They had a higher likelihood of returning yesterday than they do today.
We have used Average Deposit Amount to define the player value, but you are free to add whichever criteria you tend to use.
Segmentation Example
Segmentation Example
What to notice from the above set-up
- Change in the State of Engagement is less than 0. What this means is that there has been a negative change and that the player is less likely to return today than they were yesterday
- With this set-up, it does not matter how many segments players went down. It could have been -1 meaning going down one segment, for example, from medium to low, or -2, from High to Low.
- You will be able to reach any player who has made a deposit on a minimum of two different days as this is the requisition to be assigned into any of the seven segments.
- Besides to Probability of Return segmentation, we have used Average Deposit Amount. We've added it to assign a player into a specific value bundle, in this case, High-Value players.
Where to go from here?
We have covered the basics of how to set up segments for your Predictive Churn Prevention Activities. Now it's time to go through tips to make sure your campaigns become a success.
Planning Effective Churn Prevention
/tutorials/predictive-churn-prevention/planning-effective-churn-prevention