Spotler Activate Predictive AI is a premium app to improve audience targeting. This feature uses predictive models in collaboration with Activate Pro (former Spotler Cross Engage) to help create audiences based on predicted purchase behaviors and churn risk. It adds a new layer to the targeting strategies already possible in Spotler Activate.
The Predictive AI module is divided into two parts:
Overview
In this module you will see an overview of all results regarding Predictive AI.
Orders
Predictive AI offers a comprehensive analysis of customer purchasing patterns. By evaluating the average order value and the number of orders over the past 30 days, you can identify trends and refine your strategies accordingly.
Using data from all your purchases, Predictive AI generates forecasts to guide your decision-making. In the orders section, you can access a historical summary of your merchant activity, providing insight into the average order value over the last 30 days.
This information helps you make informed decisions about future predictions, such as determining customer value and identifying the optimal customer lifetime value (CLV) for retargeting.
CLV prediction
Here you can see the predicted average amount of orders and value per customer. You can use the dropdown to select the prediction period; 30, 60 or 90 days.
Churn prediction
With churn prediction you can see the proportions of the different churn groups. You can follow up customers which are coming into the medium churn group, or exclude people who have a low churn rate, as they will probably Purchase again.
Revenue Prediction
When looking at customer data, two key numbers can help you understand your audience:
- Number of Profiles – How many customers belong to each group (decile).
- Mean Value – The average predicted revenue per customer in that group.
These two metrics helps you decide where to focus your efforts—whether it’s marketing, customer support, or special offers.
The revenue prediction is divided in:
What are deciles?
Imagine you split all your customers into ten equal groups, ranked from the most valuable (Decile 1) to the least valuable (Decile 10). Each group contains roughly the same number of customers, but their spending potential is different.
Customers are segmented into deciles based on their predicted Customer Lifetime Value (CLV). Higher deciles (e.g., Decile 1, 2, 3) contain customers with higher spending potential, while lower deciles (e.g., Decile 9, 10) include those less likely to spend much.
- Decile 1: Represents the top 10% of customers with the highest predicted CLV. These are your most valuable customers, ideal for targeted offers or loyalty programs.
- Decile 10: Represents the bottom 10% of customers with the lowest predicted CLV. These customers may require re-engagement campaigns to retain them and increase their value.
Predictive revenue to deciles
This graph displays the total value of each decile, providing insight into how they compare to one another. Depending on the number of profiles and purchases in your merchant data, deciles 1 and 2 may clearly outperform the others, while the lower deciles may show significantly lower values. With a large dataset, decile 1 often stands out as the strongest, but its growth may quickly level off.
The highest deciles represent customers who are most likely to make a purchase. You can further increase their likelihood of buying through retargeting. Since these customers generate higher revenue, you may choose to engage them through premium, high-converting channels such as paid ads or WhatsApp.
Create audience
Here, you can see the number of profiles within each decile along with the median value (Note: Mean here refers to the middle value, not the average). The deciles are nearly equal in size.
If you plan to retarget specific deciles through ads, this view helps you determine whether the groups are large enough to be selected within ad platforms.
To create an audience for a specific decile, simply click Create Audience. This will take you to the audience builder with the appropriate conditions pre-applied.
Number of profiles
This tells you how many customers are in each decile. If one decile has a much higher number of customers, it means that group represents a larger portion of your audience.
- If high-value deciles (e.g., Decile 1, 2) have a lot of profiles, that’s great! You have a big group of strong potential buyers.
- If low-value deciles (e.g., Decile 9, 10) have a huge number of profiles, it might indicate that many of your customers are low spenders—which can be an opportunity to improve conversion strategies.
Mean Value - Why is it lower than AOV?
This shows the average predicted revenue per customer in that decile. A higher mean value means each customer in that group is expected to spend more.
However, this number is always lower than the Average Order Value (AOV) because:
- AOV only looks at completed purchases, so it assumes every visitor is already buying something.
- Mean Value accounts for all customers in the decile, including those who might not purchase at all.
- Since not everyone in a decile will convert, the predicted revenue per customer is naturally lower than the AOV.
Example calculation (with conversion data)
- AOV = € 55 (this is the average order value for those who complete a purchase).
- Mean Value for a decile = € 5 (this is the predicted revenue per customer in that decile, whether they buy or not).
We can calculate the conversion rate using this formula:
Conversion Rate = Mean Value/AOV
So €5/€55 = Conversion Rate of 9.1%
This means that only around 9.1% of the customers in this decile are actually making a purchase. The rest either don’t buy at all or haven’t been engaged effectively.
Why revenue prediction matters
By looking at these two numbers together, you can make smarter decisions about your marketing and sales efforts.
- If a high-value group has a lot of customers, consider personalized offers to keep them engaged.
- If a low-value group has a huge customer base, you may want to test strategies to increase their spending (discounts, loyalty programs, better onboarding).
- Ideally your Cost Per Acquisition is less than the CLV in each decile.
This simple breakdown helps you understand your customer potential and where to invest your time and resources for the best results.