GIS FUTURES – GIS for Business and Service Planning

After “the past and present of customized and proprietary GIS”, the authors continue with the “GIS futures” at the time, specially for retailers – both for site selection and for other tatics, like retail performance modeling and network performance modeling too.

“Given the range of issues that retailers are now addressing and the more complex problems that GIS will need to consider, proprietary system development must proceed by responding very sensitively to the needs of the client. Only in this way will development remain relevant. There will be a greater role for additional support services alongside software systems and a greater need for tailored and customized solutions for particular problems.

There is now an abundance of retail space. This can only generate greater competition amongst retailers. When things get tougher, retailers will adjust their strategies accordingly. This will lead to a greater emphasis on micro-marketing tactics, such as the adjustment of a branchโ€™s offerings in terms of style of layout, product mix, etc., to provide for the geodemographical fine-tuning techniques to ensure that the branch is maximizing its opportunity in its own micro-market. Along with this will come the development of more niche retailing, in which retailers pick a very narrow market sector and design their offering to trade in just that sector. In both of these cases, a detailed knowledge of catchment areas, and their demographics, will be absolutely vital. In other words we should anticipate a greater use of GIS tools and database management in the future.


  • Data integration

Since about 1980 a whole industry has grown up around the use geographic information in business applications. The magic word โ€œgeodemographicsโ€ was deemed in the 1980s to provide the elixir of eternal success. In fact, for a decade, the market analysis industry followed the classic product growth cycle. There are now signs that this relatively new industry is reaching maturity. It is no longer looked upon as the โ€œCinderella of marketingโ€. Many marketeers now regard geodemographic and other GIS-based analysis as an integral part of the task to maximize sales potential through accurate consumer targeting.

There are certainly gains to be made in the development of better classification systems and better software for handling geographic data.


  • Retail performance modelling

The use of graphical displays to front-end large GIS can produce very showy results. But it is often said that you can get away with murder on maps if you do not know what is in the โ€œblack boxโ€ that produces them. The inner workings of the models inside the black box are mysterious to many people. On the other hand enhanced GIS software provides a unique opportunity. For the first time, GIS makes it practical for many clients to use sophisticated modelling techniques.

For example, regression and gravity models can be harnessed to provide meaningful and focused bench-mark answers for their users, and geography and demography are major components. However, in the past, relatively few organizations have benefited from their existence. For many retailers the cost of collecting appropriate data, hiring high-level expertise and developing a model is beyond their means. However, as models become more easily configured in off-the-shelf business GIS software it is likely that they will be used more commonly within the industry. It is when researchers can access GIS software that can capture the power of different models, without having to do anything other than define or reference their data-sets, that they will recognize the value of this technology.

Geodemographics is not the only determinant of a successful site. Although the ability to discover what kinds of people are living around a store was a major leap forward, the market is now more demanding. Retailers want to be able to link geographic information to many other characteristics about a site. For example:

  1. Micro-site โ€“ position on the high street, traffic flows, pitch quality, location of nearby shopping magnets, car parking.
  2. Store details โ€“ window displays, floor-space, store appearance, design and layout.
  3. Staff details โ€“ number of staff, quality and training
  4. Competition โ€“ presence of major competitors in the micro-environment and in the larger catchment area.

Although many of these facets are not directly geographical, the need is to be able to link these with geographical data to produce more realistic potential models and turnover prediction models. The data to measure some of the above determinants are now more readily available than they were ten years ago. For example, external data suppliers have collected data on the location of major retailers which can be used for competition analysis, and computerized payroll and staff-management systems now hold data within companies to address staffing characteristics. Store details are also more readily collected, especially for merchandising purposes, and there is a greater willingness to collect other information via store surveys for turnover prediction analysis.



For retailers of goods and services distribution is the crucial component of their businesses. Their products are delivered to the consumer market through a set of distinct and discrete units: shops and supermarkets, bank branches, motor dealerships, fast-food restaurants or even schools and hospitals. Each of these units comprises part of a supply network with its own distinct geography. Investment in the distribution channel is often large as significant competitive advantage in terms of brand, format and location can generate significant profits. During the 1970s and 1980s this investment was largely associated with network expansion as the main multiple retailers, supermarket groups and financial service organizations battled it out for increasing market shares (Kay 1987). As the recessionary climate of the 1990s does not necessarily encourage such expansionary strategies, then the battle for market share (and profitability) shifts to the search for greater returns from the existing store network. For this reason, it is essential for retailers to understand the performance of their network and of individual outlets within that network: only then can they assess the potential for changes to that network and forecast and monitor the threats from competitor market strategies. To achieve this information and intelligence on what has been, what is, what if and what should be.

The philosophy of spatial modelling recognizes that there are three main components of a market system: demand for a product which is spatially variable across population groups; supply of a product through discrete locations such as shops or supermarkets; and interactions between demand locations (households or groups of households in census tracts or postal areas) and supply points.


How can we help predict what might happen if the store network changes?

Understanding the current market conditions is only the first step towards an effective SDSS โ€“ Spatial Decision Support Systems. There are a whole set of โ€œwhat ifโ€ questions which spatial models are ideally designed to address. Despite the overall trend towards consolidation of store networks a number or organizations are still committed to expansion programmes within particular market niches. For example, the UK discount retail chain โ€œKwiksaveโ€ announced in November 1993 that it intends to open the equivalent of one new store per week for the next two years. Similarly, the fast-food retailers โ€œTaco Bellโ€ in the US have a plan to open 3000 new outlets across America over the mid-to late 1990s. There is also an increasing international dimension to retail store expansion. As major home markets become increasingly saturated so retailers are looking to expand internationally. Wrigley (1993) provides example of this type of โ€œspatial switching of capitalโ€ in the grocery sector. For all these types of groups (and performance levels) the ability to forecast or predict likely revenues is crucial. The model should be able to predict not only turnover levels (by segment) but also a range of performance indicators related to market shares and profitability (Birkin et al. 1995; Birkin 1994).

Not all organizations will be interested in spatial modelling for assessing the impacts of new store openings. Manu retail businesses are looking to rationalize existing branch networks. The banking industry is a good case in point. Many of the UK banking organizations now feel they have too many branches on Britainโ€™s high streets and are looking to make significant cost reductions through branch closures (especially as ATMs and self-service banking become more common). Once again it is important to make sure that the right branches are closed and the potential damage to market shares is minimized. A third use of โ€œwhat ifโ€ capabilities is to examine the impacts of refurbishments. This may involve the realignment of product ranges to match more closely the nature of local demand. In banking, for example, it may be clear that a particular branch is performing poorly in mortgage sales given the high rate of new house construction in an area and the high rate of new household formulation (and hence persons looking for a first-time mortgage). The bank may be able to respond by increasing the support services for mortgages and shifting staff away from less productive areas. Again it is possible to insert such changes into the model and to recalculate likely revenues and effects upon the competition.”

Published by




Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s