15 Nov Happy National GIS day! What’s in your Geo-Toolbox?
Like National Cake Day, Pie Day and Pizza Day, National GIS Day gives us an opportunity to share with others what’s not only interesting to our community, but how the GIS industry is making a real difference in our world.
If you’re like me, you probably have more GIS tools and data at your disposal than Santa has elves. This comes from my many projects over many years, all requiring custom maps, models and reports for those unique projects. And if a project requires an answer that one of my applications can’t produce, I’m either left looking for another tool, or finding an external resource to complete the work.
However, as someone who has been in the GIS industry for most of my adult life, I believe that most GIS professionals do not challenge themselves enough to use the tools and data that they already possess, missing opportunities to demonstrate the value these important assets bring to their company and their clients.
Instead of fully understanding the power and depth of their applications, many GIS professionals react to a project and look for quick and easy solutions. There’s the old saying, “You can have it quick, easy, or inexpensive. Pick two.” If we’re not careful, we’ll miss an opportunity to demonstrate the depth of our GIS knowledge by deferring to additional tools and even outside resources.
Many times, we already have the solution in our toolbox – we just need to pause and think about the problem for a moment, make sure we understand the depth and power of our applications, and then apply our GIS experience to get to the answer. The result will not only be a meaningful answer to a difficult problem, but also a means to demonstrate a greater value of our GIS knowledge and GIS applications for our stakeholders.
Allow me to illustrate this with a classic GIS problem – site selection and predictive analytics.
Typically, a map “pin” and corresponding demographic and competitive analysis study area reports will be produced to show a potential site. From this information, companies will decide to either move forward with the location, or continue looking at other sites. This type of analysis is best known as “predictive analytics.” Predictive analytics applications use industry and customer data to develop a model that provides an estimate of performance of a potential site and the potential impact of the site on nearby stores.
Often predictive analytics is perceived to be the domain of GIS analytic companies who offer models that take a long time to develop, be very, very, very expensive, and, most importantly, tend to be “black box” in nature. I have found that for many site selection projects, paying big bucks for predict-o-matic models is just not necessary. If a GIS professional has the right understanding of the applications in his or her toolbox and a basic understanding of data and how to apply it, he or she can apply the assets at hand without the expense of pricey third-party models. This may not replace a seven-figure predictive system, but it often gives a low-level six-figure system a run for its money.
One approach I often share with GIS cohorts uses customer data, segmentation, MapInfo Professional, and AnySite. When combined with a bit of intent, skill and patience, these ingredients offer you the ability to produce in-house analytics-based site selection decisions – and is almost always a better bang for your buck.
The first ingredient is customer data. The more granular the data, the better the results. In the example below, you’ll find customer sales data that has been geocoded. Precision is very important, and should be at least to the ZIP+4.
Using AnySite’s Segmentation Tool, a segmentation cluster code is assigned. Adding the segmentation cluster code lets us identify the customer’s persona so that we can develop an average spend for customer household based on segmentation type. [Click here If you would like to learn more about segmentation groups and clusters.]
Next, using the current store locations, a series of drivetime captures to determine the drivetime needed to capture 75% of the sales for the existing stores are created. This establishes a 5-minute geographic benchmark as the basic trade area size of our existing stores.
Creating a map of the research area MapInfo Professional. The Grid Tool is used to create a one-half mile grid system encompassing the market area – we will focus on the Los Angeles and Orange County areas. The Grid Tool is one of many map-basic applications that is included with MapInfo Professional.
Here is the step that many GIS professionals may not know. Using AnySite’s Batch Processing, generate a five-minute drivetime boundary for each of the cells. This will allow you to generate data for estimated sales within the five-minute drivetime grid for the market area. The data is created from the average sales for each segmentation persona and the count of Households for each persona found within the drivetime.
This processing may take some time and there are tricks to make the processing go faster, but its usually best to kick it off and go grab a quick cup of coffee. And, if you are limited by computational speed, the data can always be broken down into smaller, more manageable parts and reaggregated later within MapInfo Professional if you are not able to process the entire dataset in one run.
When the Data for the grid cells is generated, the data can then be linked to the grid cell boundaries in MapInfo Professional. This creates a table of data not only available for use in AnySite and MapInfo, but can also be loaded into other systems. An example could be an RDBMS (Access or SQL) which can be handle huge sets of data. AnySite or MapInfo Professional can also access and use the data.
After the data is linked to the grid cells and added to the map, it can then be displayed in its final form. Displaying the results as a thematic map offers good data visualization.
When estimated revenue is mapped out in this manner it is commonly called a “Green Field” analysis, and can be found in even the most complex models. It provides insight to a market by identifying areas where both the location and dollar amounts of revenue are displayed on the map. Using a grid as opposed to a geography like Census Tracts makes for a smoother looking map which many find easier to visually understand and consume.
As seen above, adding existing locations and\or competitors brings even more insight to the map.
Generating five-minute drivetime boundaries around the locations will provide even more locational insight to what is going on in the market.
Of course, bringing the map back into MapInfo Professional will let the user make any final adjustments or changes to the map, making the map even more meaningful as the following examples show.
While displaying the results on a map make it easy to see the results of site research, GIS professionals must always be prepared to generate some “hard numbers” in the form of reports. It’s often easiest to leverage the grid data set work that has been visualized on the map. An Estimated Dollar Amount for a study area based on the household count and average spend for each persona is easily produced during the “site research” portion of the market development process. This uses the same report that is used in the grid data generation process.
However, what if I wanted to make the results available in common-place office tools that readily accessible by GIS and non-GIS folks alike?
One of the little know features of AnySite is that the generated reports are placed into an Excel spreadsheet template that is stored in AnySite’s Template Directory. Reports created within AnySite can be set up to “dump” the data of a report into a custom Excel document that can be manipulated by any experienced Excel user. With this in mind I offer the following.
Since drivetimes are a function of sales, I will assume that the distance from the site to the customer’s household directly affects customer spend. With the right use of tools, this opens a whole new opportunity for the GIS professional to enter into the more “predictive” aspect of site selection and market analysis.
A report in AnySite collects the data within a study area around a potential site. Using a custom excel template allows the information to be used to predict sales for a potential location. The report can use the distance of the geographies from the potential location and modify their contribution to the overall projected earnings along a decay curve that is based on the distance from the potential site. The result offers more “predictability” of the potential site, rather than considering basic demographic data alone.
Don’t lose site of the fact that the key tool here is Excel. The resulting spreadsheets are of your own creation, allowing you to modify the parameters of the analysis according to your business requirements. Site configuration and regionality may influence the shape and distance of the decay curve you used. And because you’re using tools at your disposal, you have the added flexibility of easily analyzing current locations relative to competitors and how they might be taking a piece of the “sales pie.”
It is these decisions specific to your company that makes your job as the GIS profession indispensable and more valuable to your company. While someone may argue a multi-million dollar modeling system may be better, it also will most likely have a significant “black box” feel with components that are out of your control to modify and expensive if changes are necessary.
Oh….and let’s not forget that it can take those larger models months to be ready to produce results. The example I used above took less than a week, and most of that time consisted of computer crunching the data – while I worked on other projects.
So, on this GIS Day, take a moment and think about all the applications that you have in your toolbox and how they might work together. Don’t be afraid to explore. You may just discover paradigm changing methodologies that improves your site selection process. You may be surprised at what you can do!
If you have any questions or would like to learn more, click here and let me know. I’d be glad to talk with you more about how to get the most out of what’s in your GIS toolbox.