gmcgarryWritten by Ian Mc Garry, CMX Software Engineer

The release of CMX 10.2 is coming soon and with introduces the improved Correlation Widget. Allowing you to view the relationship between areas, Correlation is a powerful tool. To help introduce you to the widget we’ve put together some simple but beneficial use cases outlining its utility and power.

Correlation Use Case – Identifying the Relationship Between Areas

Certain areas within a location, such as shops, should have a natural interaction or correlation with each other. This can be due to their related nature, ease of accessibility or even close proximity with one another. For example in a mall we may have a flower shop and a card shop. A strong correlation between these shops is expected due to the well-known fact that flowers and gift cards go hand-in-hand. Using the Correlation Widget we can quantify this relationship and prove or even disprove our theory.

We set up a simple correlation widget that looks at the data for shops last week. Setting our Card Shop as the focus allows us to then see the Correlation between it and the other shops in the building.


From this chart we can see that there is a 25% Correlation between the Card Shop and the Flower Shop. This tells us that out of the 100 Visits that occurred in Card Shop, Last Week, 25% of these Visits also happened to visit the Flower Shop. This may be lower than expected especially when we compare it with the higher correlation between Card and Cake Shop. This may be due to proximity or better signage. We can now go about modifying our shopping center environment to better direct visitors from the Card Shop to the Flower Shop. Upon further investigation we find that the signage directing visitors is insufficient. After correcting this we can then check if there has been an impact. To achieve this we change the data period to This Week allowing us to view the change.


Here we can see that the Correlation has increased significantly from 25% to 50%. This can be attributed to the changes made to signage around.

Correlation Use Case – Identifying Popular or Trending Stores

Some stores within a location, such as a Mall, will tend to control a more prominent visitation when compared to others. This may be attributed to their proximity to the entrance, their importance to customers or just their ability to draw customers in. CMX Analytics is well equipped to answer the question of “How popular is a store when compared with others?”. Using the correlation chart we can observe and quantify these relationships.

We set up a simple correlation widget that looks at the data for shops last week. Setting our Main Entrance as the focus allows us to then see the Correlation between it and the other stores in the building.


From this we can see that Grocery Store is probably the most popular and the most visited store in our Mall by people entering via the Main Entrance. Again this can be attributed to proximity, popularity, signage, ease of access and a variety of other reasons but the chart quantifies the relationship for us so that we can come to a natural conclusion. In this case we attribute the result to the Grocery Stores necessity amongst visitors to the Mall as nearly everyone needs to shop for Groceries. Coming in at second is the Coffee Shop which also makes sense as we usually say that people who go shopping like to sit down for a coffee afterwards to relax.

We hope you enjoy working with this new feature of CMX 10.2 and look forward to hearing how you have been putting it to use in your location!


Daryl Coon

Cisco Customer Solutions Marketing