Location Intelligence is a branch of analytics that uses geospatial data to uncover insights and answer business questions. Retail is a great example of an industry that can benefit greatly from location intelligence, because it has a significant geospatial component. Brick-and-mortar stores are physically tied to a location, and all customers originate from a location, such as their homes or their workplaces. If a retailer wants to open a store, they want to locate it somewhere that customers can easily find. Location intelligence can help them determine the optimal location.
One of the things we look at in location intelligence is the relationship between retail locations within a specific area. Whether they’re physically right next to each other or are part of a shopping centre or plaza, they can impact one another as part of a grouping or cluster. A cluster represents a single location of combined attraction. Customers come for one store and then have a high likelihood of also going to the other stores in the cluster, given their close proximity. We see clustering even in retail locations that are physically distinct (across the road or down the street) from one another. In such cases, the cluster represents a retail area, such as a strip or an intersection, rather than a singular location.
We create a cluster by first determining which retail locations we want to analyze, usually within certain subsets of stores, such as specific retail categories, different banners that belong to the same parent company, or even retail locations that all use the same loyalty program. Then we calculate the distance between each of the stores; those that are a short distance from each other are considered to be part of the cluster. Once we’ve identified the retail locations that belong to the group, we can plot the cluster on a map, either using the point at the centre (centroid) of all the locations or the average of their coordinates. The cluster itself does not necessarily have a specific geographical location; rather it’s a point on a map that approximately represents a concentration of retail activity.
Why use clusters?
From a location-intelligence perspective, having locations and customers clustered into points on a map makes it easy to calculate the distance between them. Knowing the distance is important to understanding how far away customers are and how far a retailer needs to extend their flyer mailings or roadside advertisements. Grouping locations into clusters also makes it easy to visualize concentrations of retail activity by looking at the distribution and density of clusters across a larger market.
Another benefit of this method is that it allows us to look at the retail mix or combination of locations that make up each cluster and identify whether any patterns exist. This can be very helpful when trying to understand a market, as well as in determining where to open a new store and what kind of store to open. Knowing where a larger cluster is located, or where multiple clusters are concentrated, can also help with predicting traffic or activity levels within an area. A larger cluster consisting of many retailers can be very attractive and can drive significant activity to the area.
We also create clusters for competitor locations and use them to identify areas to avoid (e.g., where there’s a strong competitor presence or a larger cluster in the area). We can conduct other analyses to understand the impact of clusters in close proximity to each other.
Location intelligence specifically looks at the geographical or spatial aspects of data analytics. This spatial component – where things are – is what makes retail clustering possible. Grouping retail locations together is just one of the many ways to analyze spatial data, providing a different and supplementary view of a market. It has the potential to uncover new insights that are not otherwise apparent, and can help show retailers where to open their new stores.