Using Geographic Information Systems: A Different Perspective on Data

Bookmark and Share

Image courtesy of Shutterstock

Geographic Information Systems (GIS) technology allows users to engage with large, relational datasets in a map format. GIS adds another dimension of understanding to traditional databases through its emphasis on location. This technology significantly enhances business intelligence in three key ways: data integration, visualization, and spatial analysis.

The label “GIS” can be applied to any computer system that stores, displays, and manipulates geographic or spatial information. Spatial analyses have been employed as problem-solving techniques long before the digital age. Today, GIS technology can be applied to a wide variety of situations, from township zoning to cell phones and GPS units to international crises. In 1999, the third Wednesday in November was set aside to mark International GIS Day, a celebration of GIS technology, its users, and its applications.

How GIS Works

GIS allows users to operate at any scale, from single neighborhoods to the entire globe, making it a useful analysis tool for any sized business. These computer systems are able to store geographic data in vector formats. This type of data is geometric; it maintains spatial information as points, lines, and shapes (or polygons). Vector data can be used to identify “attributes” (or map features) including specific locations, road networks, and political boundaries. Groups of related attributes are stored together in a single “layer.” For example, all road line segments comprise a single layer that is separate from other attributes such as business locations or county boundaries. By stacking multiple layers onto a single display, GIS can reveal more thorough and nuanced business trends than a dashboard full of one or two variable charts and graphs.

Over the last 20 years, businesses and industries have seen an exponential growth in digital resources, including the storage and transmission of large data sets. Even moderately sized businesses can afford to implement off-the-shelf customer relationship management (CRM) software, which documents not just sales but customer details and contact records as well. All CRMs frequently record geographic information, including IP locations, customer addresses, or address parts (e.g., ZIP codes).

A company’s CRM system is the first and most valuable data repository, but it is not the only one available. Many government agencies and international organizations publish open-source data, which is freely available to any individual or entity. A national census can provide essential information regarding demographic patterns, while World Bank data can speak to the distribution of wealth across a continent. If the necessary data sets aren’t freely available, there are many research and aggregation firms that can pull data for a nominal fee.

When combining multiple data sources, assigning a common indexing key for joins is an early but critical requirement as different data providers may conceptualize the same field in different ways. GIS can combine a variety of data sources—client-owned, open-source, and paid research—without direct links among the data sets. While traditional databases rely on key fields to establish and maintain relationships, geo-databases can also use spatial proximity as a method of linking objects to one another. Proximity does not require a matching key, only that each database is connected to some kind of geography (e.g., states, counties or latitude/longitude coordinates).

Objects from each database, mostly tables and queries, can be displayed on the same map, even if they don't reference the same geography. Displaying multiple layers at once can identify trends, explain why those trends are occurring, and determine how long those trends have been in place. Maps are different from traditional analysis visualizations (i.e., charts, graphs, etc.) because they are able to display several large data sets over both space and time.

GIS analysis capabilities don’t end with mapping displays. Most GIS software also includes toolboxes for analysis and predictive modeling. Every business unit’s budget is finite—using spatial analysis helps teams work more efficiently to ensure intelligent decision-making rooted in reliable data.

While the analytical toolbox is typically diverse and adaptable, here are three examples of improved client outcomes:

Consumer Location Assessment

Any corporate communication, whether internal or external, depends on connecting the best messages with the appropriate audience. The GIS toolbox can help marketers find their audience and target them specifically, minimizing cost and maximizing audience attention.

A regional entertainment attraction for families with young children was interested in generating a regional admission coupon across three states representing their key audience—New York, New Jersey, and Pennsylvania. With a limited budget prohibiting complete coverage across the three states, the client employed both owned and open- source data to prioritize the best areas for coupon distribution.

According to previous seasons’ attendance records, approximately 80% of visitors reported home zip codes within a 90-mile radius of the client’s sole venue. The client already had other media investments in the market, including radio, print, and television, and decided that the coupon should circulate in areas where other media coverage was light. Zip codes with complete coverage from more than one media outlet were removed from consideration. These two parameters created a “donut” of opportunity. The remaining zip codes were within 90 miles of the venue, but tended to be outside of a 30- mile radius, where media coverage was heaviest.

As a family-friendly arts-focused attraction, the client’s target audience is mainly comprised of mothers with young children (ages 3-9) or older adults who might have grandchildren in the target age range. It was recommended that the client distribute coupons more heavily in areas that had high populations of women, families, and/or older adults. ZIP codes with high percentages of the target populations were identified using open-source U.S. Census and American Community Survey data and then ranked according to statistical significance.

To test the efficacy of the areas selected, the client recorded the number of coupons returned in conjunction to the visiting party’s zip code. As part of the post-analysis, the number of coupons sent to each zip code was compared to the number redeemed to calculate the rate of return. To optimize couponing and other marketing initiatives planned for the following year, these calculations were added to a map identifying areas which responded well and which didn’t.

Selecting a New Branch Location

A business’s physical locations in the marketplace have a dramatic effect on the types of customers they attract. GIS has several tools that assist in situating new branches in areas that are easily accessible to the desired consumers.

A financial clients, a regional bank, serves both B2C consumers and B2B investors. While they already had three branches in the region, they were looking to open another. Approximately 85% of visitors to branch locations were B2C clients, but B2B investments accounted for nearly 65% of the bank’s total business.

Because B2B clients represented such a significant percentage of business, the client wanted to prioritize access from other companies in the region. This had the added benefit of allowing the employees of those companies easy access before or after work as well as during the lunch hour. The North American Industry Classification System, a limited access data set produced by the U.S. Census Bureau, identified the locations and approximate sizes of business in the region. Areas with a high density of businesses above a certain size were given first priority for potential selection.

Mapping the home addresses of current B2C customers against American Community Survey data determined that the average median household income for bank patrons was greater than $45,000 per year. Using census tracts, one of the smallest geographic unit published by the U.S. Census Bureau, Klunk & Millan was able to precisely identify areas where the population matched this criterion.

According to customer addresses, more than 60% of customers lived within a 20 minute drive of the nearest location. When selecting the location of the newest branch, all areas within that drive time to pre-existing branches were excluded.

As the new facility was opened, Klunk & Millan worked closely with the client on outreach to qualified, nearby businesses and consumers. The client continued to record customer information, especially home or business addresses. In the post-analysis, this data was displayed on a map to confirm that this location followed the same visitor and business trends that the older branches displayed.

Managing Sales Territories

Sales and service territories must maintain a delicate balance of workload and sales potential while also maximizing efficiency. GIS can be used as a predictive modeling tool to delineate the most appropriate geographic regions for a business.

A rapidly growing industrial manufacturer produces devices used extensively in the power utilities and oil/gas industries. This company wanted to organize their expanding sales team with territories that balanced the workflow of current customer service and nurtured future opportunities. They used GIS to establish their nationwide sales territories.

The first step in this process was an analysis of the company’s recent sales, particularly the customer’s billing address and their purchase history. Customers were grouped by county according to their address. Open source data from the Bureau of Labor Statistics (BLS) provided counts of employees by industry and employment projections for all counties in the United States. An algorithm that appropriately weighed current customers’ purchase history was developed that populated recent industry counts and projected future growth. As a growing company with an expanding customer base, this client prioritized cultivating opportunities. Each county in the United States was assigned a “score” which weighted the volume of both current customers and future prospects.

These scores were displayed on a map to show the client where they had already attracted a large number of customers and where there was potential for more growth. Every county across the nation was grouped together so that the sum of their scores fell into approximately the same range. No one territory contained significantly more customers or prospects than another, creating a balanced workflow for every member of the sales team.

Team members were also provided with additional insights regarding their assigned territories. Some regions contained more potential prospects than current customers, while others contained more oil/gas than power utilities consumers. Even within a region, certain cities and counties showed more promise than others.

To test the efficacy of the new sales territories, the client was assisted in instituting a robust CRM system that recorded information pertaining to both lead generation and sales. Of particular interest was the contact’s billing address, industry (power utility or oil/gas), and their progress toward or completion of purchases. In the first year post-analysis, this information was displayed on a map to ensure that the distribution of opportunities was balanced across territories. In under-performing areas, regional managers were assisted in refining messaging that better connected with the industrial composition of the territory.

Enhanced Business Intelligence

GIS and geospatial analysis techniques are inherently flexible and can be customized to suit any company. In addition, analyses can be carefully tailored to any budget, geographic region, or set of goals. GIS technology allows users to engage with large, relational data sets in a map format to add another dimension of understanding to traditional databases through emphasis on location. When used appropriately, this technology can significantly enhance business intelligence.