eRFM Visitor Module

A key feature of the eRFM audiences, is the unique ability to optimize audiences by automatically generating multiple duplicates of the selected eRFM audiences (with slight changes in recency frequency and monetary value). Once the eRFM visitor audience has been created, it will begin to optimize for up to 24 hours and then the Madgicx algorithm will choose the most optimized audience to lookalike from.

How Does it Work?

Let's take Frequent High AOV Potential Visitor Audience as an example:

  • The software will start by creating an audience based on the most broad parameters, visitors.
  • From there, it will continuously optimize to get more specific, as long as the audience stays large enough.
  • From visitors, it may then optimize from the first audience to create a second audience of visitors who have visited multiple times or have spent a significant amount of time on your page.
  • From audience two, if it is able to, it will then create a third even more specific audience. For example, it may optimize for an audience of visitors who have visited multiple times AND have spent a significant amount of time on your page. 
  • The system will continue to optimize to get as specific as possible, until the audience is no longer large enough to deliver.
  • For future reference, all audiences will be labeled as Optimized, so that you can find them later.

AOV- Average Order Value 

This Module uses information from your pixel to determine lookalike audiences based on visitors to your site. These audiences will be broken down into 18 different audience types.

The 18 audience types will be combinations of the following 3 aspects.

  1. Low, Medium, and High AOV- the software breaks down the average order value to three levels low, medium and high, by dynamically calculating the prices on your shop.
  2. Frequency - The software will break down the number of visits of website visitors into three levels, infrequent, occasional, and frequent
  3. Intent- The software will break down your visitors behavior into two levels, High intent, meaning the visitor took an action that indicates his intention to purchase, and mid intent, which refers to regular visitors who did not show any specific intention to purchase.

**Frequent, Occasional, or Infrequent Audiences represent the quantity in which the user performed a selected engagement.


There are visitors who are more valuable to your business than others. Some visitors leave after a couple of seconds, whereas others deeply engage with your page and show high intent to perform the desired action. 

Using the Frequent High AOV Potential Visitor Audience, you can build a segment of your most frequent visitors who demonstrated the highest intent to convert. 

These audiences tend to perform significantly better since the custom data we use to build the look-alike audience is made up of visitors with high quality traits, and due to the unique optimization process which takes place outside of Facebook's platform.

If the algorithm can create a fully optimized audience by adding all of the rules then the algorithm will automatically apply an “Optimized” label to the audience name within the Ads manager. Leverage AI technology to discover and target your most profitable audiences. Launch audiences from the eRFM Customer Module for a complete setup of the eRFM audiences.


Did this answer your question?