The Visual Crossing Location Analysis Designer is a rich, interactive tool for analyzing the potential performance of a set of locations and the impact of changing those locations. For example, if you are a store owner, the tool will help you answer:
- “How many potential customers do I have based on my store location?”
- “How do my competitor locations influence my store performance?”
- “What is the best place for a new store location?”
- “What stores can I remove while minimizing the impact to my business?”
The Designer first asks what goal you would like to achieve. There are three possible goals:
Analyze Current Business - allows you to investigate the potential performance of your current stores with respect to a set of demographic information (such as total population or population by age group).
Add Business Location – allows you to investigate the impact of adding a new location to your locations. Areas of information include the overall increase in possible population access and the individual change and cannibalization of your current locations and competitors.
Find Optimal Location – allows you to identify the location that would result in the largest increase in potential performance. For example, a store location that would result in the highest customer population increase.
Analyze Current Business
To start an analysis, open the editor and select the goal ‘Analyze Current Business’
After we have selected a goal, we then need to specify our current locations and any competitors that we would like to include in our analysis. In this first example, we are going use a set of sample store and competitor locations that are included in the designer. Click ‘Use Sample Data’ to use these locations.
The tool then loads the sample locations and calculates the initial analysis. This includes using the default demographic measure of ‘Total Population’ and populating a map displaying the probability distribution of visiting one of your store locations.
In the result, you can see that the total population reach of your stores is 731,600. The map has then color coded the area to indicate the probability of a potential customer at a location visiting one of your stores versus a competitor store. Green and blue colors indicate that most people in that location will visit your store locations. Reds and oranges indicate that most people will visit a competitor location.
We will discuss the model in detail later but in general potential customers visit their closest store or the store which is most ‘attractive’. Store attraction is a measure that indicates the favorability of a particular store compared to others.
So far, we have analyzed the potential of our locations based on Total Population but many other variables are available. In the sample application you can choose between a number of core demographic variables depending on your global location:
In addition to seeing the impact to the number of people that may visit your stores, you can also estimate the revenue potential.
Click on the ‘Estimate value impact’.
Given your knowledge of the potential penetration of your store and the spending per customer, you can see the change in population reflected as an increase in revenue. Here we have estimated a 2% penetration with a revenue per person of $20. This results in a potential value increase of $292, 643.
Add Business Location
The next goal of the tool is to analyze the impact of adding a new store location. Start the tool by selecting ‘Add Business Location’
We will again use the sample data as in the previous example. We are now ready to specify a new location.
Click on ‘Add Locations’
You can add locations either by importing a file of new locations or selecting new locations on the map. Removing a location is not available in the sample application.
Click on ‘Add’ to add directly from the map.
You can specify the new location based on address or by clicking on a map location:
After clicking a location, or entering the address, the map updates to indicate the new location:
Note that you can specify the attractiveness of the new location. To help provide an attractiveness that is of the correct order of magnitude compared to the existing locations and competitors, the mean values for existing locations is used by default. Click ‘i’ for more information.
In addition, to the map reflecting the new location choice, you will also notice that the metrics for the location have been updated to highlight the change that this location causes to your potential performance. You can see that this proposed location adds a potential 4,820 people to your store customers.
If you press OK, you can add further proposed locations or add penetration and spend information to predict increases in sales. See section one for more information.
Find Optimal Locations
The final option available in the sample Designer is to add an optimal location. This finds the next proposed location that will increase the selected demographic measure by the greatest amount.
Select the option and select the sample locations
We now select ‘Find locations’ and the Designer will search the visible area for the best performing location.
This will be selected a potential store and the designer will display the results as per the previous section.
Publish the Results
For each of the above applications we can publish the raw results to an OData datasource so that they may be used as a datasource for an SAP Analytics Cloud Story. After hitting publish you will be presented with an OData URL that can be used to import data into SAP Analytics Cloud.
Attractiveness and Model
The Location Analysis Designer using a probability allocation model based on the ‘Huff Gravity Model’. This model assigns the likelihood of a set of people at a particular location visiting a particular store based on the distance to the store and the attraction of a store. So for example a customer may visit a store further than another store if it has a larger square footage, additional departments etc. The attraction of each location can be set during the import or addition of a location. It may also be modified by viewing the existing locations:
The full list of locations is then displayed:
The attraction of each item can be modified from this editor. In addition, the color of the location set can be modified which will reflect in the map color theming.
Advanced Model Parameters
You can also modify the parameters of the model itself. Increasing the Huff Alpha increases the sensitivity to the attractiveness of each store. Huff beta increases the effect of distance and so a more negative value will cause further stores to be less likely to be chosen by the consumer. Maximum distance is the maximum distance that the customer will travel to any store and serves to define the bounding area of the analysis.
Importing Your Own Data
In addition to using the sample data you can use your own data in the Designer as well. To import this information into any step, you can either click on the map to add locations (or search by address) or import a text file. To import a text file, select the type of locations you would like to import:
Hit ‘Add Locations’ and then ‘Import’
You can then import a file from a text file by selecting a file:
Note that there are structure requirements for uploaded datasets. The data set must be a text based file delimited by either tab, comma or semi colon. The first line of the file must be the column headers.
In addition, the file must include a column for ‘name’ plus the ‘longitude’ and ‘latitude’ of the locations. The attraction of each location can be optionally included (or they can be added later) using the ‘attraction’ column.
Here is an example of a correctly formatted file:
Quick Mart 1,38.923961,-77.237349,10
Quick Mart 2,38.903814,-77.26284,6
Quick Mart 3,38.901021,-77.224215,7
Quick Mart 4,38.92222,-77.201018,5