Pitney Bowes
MapInfo Pro™ Monthly Journal
Using MapInfo Pro to Visualize Analytic Processes and Outcomes

Editors note: This may seem familiar. A few of you may have read it in March. I decided to re-run it because I think it is a really good article and it was buried in the March issue. Tom Probert, Editor.

Tom Visualizing data can be tough – helping people understand what they’re seeing, making sure they understand the results and making sure they know the WHY behind the analysis. No one knows this better than those who interact with data all day and who have to explain that data to someone who doesn’t view it with the same spatial approach.

This particular challenge is the case for my team at Pitney Bowes. Day in and day out the Applied Analytics team works with client data and workflows to help our clients make good business decisions based on complex, multi-dimensional data models. However, explaining the analytical-WHY is sometimes tougher than building the predictive models themselves!

Analytics and “The Curve”

My team evaluates many aspects of consumer behavior and spend potential for our retail and restaurant clients. We help them to create sales-forecasting models. These models enable clients to choose their next-best locations with minimal risk and with maximum potential and best fit to strategy.

Part of our predictive analysis includes generating sales penetration curves that depicts consumer potential based on a customer’s proximity to a unit (store or site). Conceptually, clients understand how the curves are determined and how they will be used. However, during a presentation of these findings, there can be an “eerie silence” as they start to process these important analytic findings.

Evading the eerie silence can be difficult and reviewing complex data models can be too much or too detailed for some audiences. There may be audience members in your business that don’t need or want to know every detail of an analysis, but they do need to understand the findings in order to create the best business decisions based on the data.

the curve
Figure 1: Example: Customer spend by distance model in standard curve chart format.

A More Descriptive Visual – Maps

So, how can data be presented to show findings better and to rationalize business decisions, while still keeping the data’s integrity?

The Pitney Bowes Applied Analytics and Marketing Services team uses MapInfo Pro daily on behalf of our clients. Loading model analyses and products into MapInfo Pro enables the team to easily visualize the data modelling, decisions and findings. Communicating complex results in an easy-to-understand map adds value to the data while providing a visual, geospatial picture of it.

Evaluating Model Products in MapInfo Pro

So, the graphed curves don’t always cut it – but a map always makes the data easy to explain. Here are the building blocks for the above curve (Figure 1) in MapInfo Pro.

Building Block to Visual #1: Get the audience to the basic elements of the curve, geographically.
  • Neighborhood = Tract
  • Distance = .5 Mile Radii around unit
  • Neighborhood Densities = which Curve will be referenced
block 1
Figure 2: Building block 1 – geographic curve

Building Block to Visual #2: Relate the map elements to curves. Each Neighborhood\Tract is assigned a value directly tied to the curves based on density and distance to unit. The points (red triangles) reflect the value where distance curves intersect with customers based on local density.

wireless telecom
Figure 3: Building block 2 – map elements in relation to curve

Building Block to visual #3: Populate Neighborhoods with Customer Potential - Based on potential customers per 1,000 persons and neighborhood population, demonstrate total customers from each neighborhood.

block 3
Figure 4: Building Block 3: Populate Neighborhoods with Customer Potential

Final Output: Once the audience understands the building blocks and how the curves operate geospatially, Pitney Bowes can easily pick up the conversation to creating representations of sales distributions for all existing units or potential sites.

final output
Figure 5: Final Output Customer spend by distance model with all building block layers

Ultimately a range shaded thematic map that shows customer interaction by distance from a unit can simplify the complex nature of the sales curves and provide the audience with a visually appealing, simple to understand graphic upon which to base the conversation. No eerie silence need intrude!

Article by Tom Coats, Principal Consultant, Applied Analytics and Marketing Services

Tom, a Client Manager and Project Manager in PB’s Applied Analytics and Marketing Services group for over 10 years, has worked with retail and restaurant industry clients to provide insight and solutions in the areas of real estate, franchise development, and marketing. When not working on his historic home in Toledo, Ohio he can be found annoying his wife and cats with awful guitar covers of 1990’s power-pop standards.

About PB’s Applied Analytics and Marketing Services:

Pitney Bowes’ Applied Analytics and Marketing Services team analyze spatial and descriptive data to power and predict smart business strategies. Our expert consultants, better data, and leading-edge software help you locate and act on opportunities faster than your competition, transform delivery channels with confidence, and market effectively based on rich customer insights. We create reliable and effective solutions tailored precisely to your business.

Questions or suggestions?

To learn more about our team, how they help clients with business decisions, or how you can use MapInfo Pro to talk with your teams, please contact Michelle Scoggins or visit our Professional Services team on the web.