Make Smarter Decisions with Geoanalytics

By Marco Brugna, Mon 06 March 2017, in category Business intelligence

ESRI, Maps, Qliksense

  
@

Geoanalytics is the approach of applying spatial analysis to our data in order to understand our world, where things are, how they relate together, and what actions to take.

In this article we’ll demonstrate how geoanalytics can help generate smart insights during a decision process. In our sample we’ve used five leading coffee shop brands located within the London area, and analysed the surroundings and competitive pressure, then defined a possible new store location.

First, we make some assumptions:

Our journey into the London coffee shops starts with mapping the data. The London DataStore is a great resource of different demographic and economic data. By combining this data with our store locations, we can start clustering the stores based on their locations and surroundings.

image1

To understand the type of store we need to define the area of influence of each location. So we started by creating what’s called a Service Area.

We decided that each store could attract people from a maximum of 10 minutes walk away. Using the newly created service areas and the defined spatial relationship, we can derive how much of the census blocks fall within the service area, and therefore how many people live within the 10 minute limit of our store.

image2

Combining the service areas and the census data, now we are able to understand the context of each store. If they have more residential or business, the level of instruction and employment, crime rate and average economic figures.

By comparing the data inside the service areas in our sample, we notice that of the 390 Costa stores, 73% of them are on residential areas, compared to only 17% of Pret A Mangers, who seem to be concentrating mostly on business areas.

image3

The next step of our analysis is to understand the competitive pressure of our brand.

Combining service area and distances, we start now clustering the stores, not only by demographic data, but also by competitive presence. We can calculate how many competitors are within the 10 minutes from each location and which are the closest.

What come’s out is that the borough City Of London is a complex battleground with an average of 39 competitors within 10 minutes walking time, with a possibility for a customer to get a coffee every 1 minute. So you really need to stand out to attract customers.

image4

If we combine sales and competitive pressure, customer data, demographic and economic data we can now apply the concept of Ansoff Matrix to define a strategy for our stores based on the location.

We could change our product line to have a take-away lunch based on the fact that it’s a business area, or enhance the coffee area if it’s closer to a Tube station in a residential area.

image5

Population trend from 2010 to 2020

Another strategy is to increase our market coverage, and we can do that by analysing different indicators. A basic strategy is to look at the current picture of your data, competitor density, actual population and find a place that maximize the visits to your new store.

Combining geo-statistical models and time-series forecasting you can predict location-based events and define your development strategy to maximise your success.

There are a large number of analyses that we can do by combining business and geographic data, and sometimes the result doesn’t need to be an output map. What is certainly sure however is that we need to add the geography as a fundamental dimension by which we look into our data.

If this took your interest, why don't you join our geospatial analytics event on March 21st at the Sky Garden in the City of London. More information can be found here.

image6

The data are property of their respective owners. We have used Esri for the GeoAnalytics elaborations and then Qlik Sense for data exploration and analysis.