As part of our initiative to provide strategic solutions for business, CARTO is thrilled to announce Gravity Models—a function that users can add to their analysis–that predicts the amount of people who will come to your location based on the presence of competing locations.

The first Gravity Model, for retail analysis, is based on Reilly’s Economic Law of Retail Gravitation, which suggests that customers travel longer distances to larger retail centers the higher the level of attraction the shopping center has. Attractiveness, is analogous to size (mass) in the physical law of gravity. Based on Reilly’s Law, businesses can make informed decisions about their next move. Imagine, for example, that one of your competitors has a store in city X, and that this business enjoys a position of market dominance in the region. If your company opens its own store in neighboring city Y, will this new competition attract enough traffic away from city X to turn a profit?

Another approach that can be applied is Huff’s Law, which takes multiple attributes and retail centers into consideration. This model’s approach to market penetration assumes that there is a spatial variation in the proportion of households served by a store due to competition. The trade area is conceptualized as a probability surface that represents the likelihood of customer patronage.

The idea is to create a probability surface that is based on a spatial interaction model that includes variables such as distance, attractiveness, and competition. The probability surface can be contoured to produce regions of patronage probability, which can then be further used as weights in the preparation of a market profile.

SEE THE POWER OF LOCATION INTELLIGENCE

Request a live demo

In this visualization you can see various mall locales in Madrid. Simply choose a mall from the list and begin your analysis! We’ve selected Xanadú shopping center, in this example, to determine how many customers are attracted to shopping in this particular mall and their distribution.

The dashboard histogram widgets update in real-time to reflect population density, potential customers, probability of patronage, distance to mall, and average age. The color ramp is dynamic, which gives you a color scale and range that best reflects your insights. Simply adjust the probability of patronage histogram to accurately reflect your minimum and maximum values denoting the meaning of your color choice.

For example, if you adjust the probability of patronage histogram to reflect approximately 25-50%, you’ll notice that the Xanadú shopping center’s surrounding areas remain a yellow indicating their high probability of patronage. The surrounding regions that are deep purple have the lowest probability of mall visits. You’ll also notice that the total population is a little over 19,000 with more than 6,600 potential customers, with most of the consumers located directly around the shopping center.

Now imagine you are a commercial real-estate speculator and you want to determine where your next mall should be, how big you should make it, and the best ROI in terms of attracting people. Select the new mall function on the right and and then select a location on the map of where you’d like your new mall to be, the cursor places a dot (colored based on the size of your mall). You can adjust the attributes in your dashboard to reflect how many square meters you’d like to make your new shopping center and start performing your strategic analysis. You can also drag your dot to a different location and begin recalculating and adjusting variables.

The function behind all the color craziness is CDB_Gravity, which you can find in this repo. This is a simplified version of Huff’s Law that can be used with different parameters, such as weights or population.

Gravity Models is a powerful way to make decisions and solve problems that can be applied to various sectors including migration trends, trade flows, transportation demand forecasting, commercial and tourism attractions, and trade areas, and CARTO is already offering the solution!

Happy data mapping!