Geography can be a very important factor in business, as direct marketers are well aware. The use of postcode geography to target and profile customers has long paid dividends. Putting a separate value on the role of geography is much harder. Yet it is increasingly important as a result of the growing role of geographical information systems (GIS).
While marketing departments may already have customer databases and targeting systems in place, GIS appears to offer not just new functions, but potentially a new business platform. Most often understood for its mapping capabilities, GIS may become a common IT currency which allows marketing to interact with site planning, distribution, salesforce and external media. Using geography as the universal reference point, data from all these areas can be analysed and acted on.
But how can the cost of building or outsourcing a GIS be justified? In many cases, the system sits in the middle of a planning function and is not always seen to affect end results. But this is beginning to change as more sectors adopt digital mapping and marketing brings it closer to the sharp end of business.
“It allows you to bring geography into play on any targeting model you use,” says John Frood, senior data planning consultant at WWAV Rapp Collins. He notes that the inclusion of spatial data in targeting models is relatively recent, yet some sectors have been gaining advantage from it for years. Automotive manufacturers, for example, have long employed GIS for dealer location.
Now these manufacturers are pushing the system into examining purchasing behaviour and loyalty. This is being driven by the recognition that if customers will only drive a certain distance to buy a car – what Frood describes as the “pull factor” of a dealership – they will also be affected by distance when it comes to looking after that car.
“Servicing and after sales generate the majority of the money for car companies. Applying spatial analysis to potential customer retention is very useful,” he says. After three or four years, a car owner is more likely to take a car to a non-manufacturer outlet for a service. By analysing this propensity against the catchment area of the dealer network, potential risk zones can be identified.
One of the key areas in which GIS can have an impact on the bottom line is door-to-door distribution. Geography is highly visible in this medium and is susceptible to careful planning. Managing director of the Leaflet Company Mark Young says: “We know by postcode sector how well items are distributed. You might say when selecting sectors using targeting systems that they would generate good business. But if they are in the middle of nowhere and can’t be easily delivered, there is no point including them.”
By using Stepcheck research to verify delivery of items, Young’s company has been able to rank every postcode sector of 2,500 houses by the efficiency of distribution. This can then be used as a filter against any selections produced by using geodemographic targeting systems to identify prospect areas.
Furniture retailer Courts has taken this one step further by including what marketing manager Colin Keyward calls “real world factors” in its GIS. From delivery information, it has 95 per cent of its customers’ names and addresses. Over the past seven years, the company has proven the value of profiling its distributions using the Mosaic profile by testing sample areas against a control. In the original test, the 15 stores which had leaflets dropped in profiled postcode sectors saw incremental sales of 1.5m.
It has now gone one step further. “We have overlaid our sales data on individual store catchment areas to find the most important target customers,” explains Keyward. Business rules are applied to ensure that top sales areas get full distribution, then postcode sectors are ranked by their Mosaic profile, before being selected against sales potential. The Leaflet Company then screens these sectors using its efficiency rankings. Return on investment can be justified from savings on printing and distribution of the 86-page Courts catalogue alone, before sales uplift is taken into account.
Gillette used GIS to identify targets for a sampling exercise carried out by Circular Distributors for its SensorExcel product. Recognising that 50 per cent of all male wet shavers already use a Gillette product, it needed to ensure that samples went to non-users. With a goal of sampling 40 per cent of the universe of disposable razor users in order to convert 25 per cent to SensorExcel, it put together a targeting brief using GIS.
The system combined catchment area data on stores which stocked the product with profiles of typical disposable users who had been identified within the local population. This gave a set of target postcode sectors to which samples were then delivered. (The sampling mechanism itself helped to filter potential users. Households received a bag on day one which they were invited to hang on the letterbox the next day if they wanted to receive a SensorExcel disposable. On day two, bags were then filled with the sample).
The distribution proved to be highly efficient, with 70 per cent of samples going to disposable razor users. Of these, 92 per cent were non Gillette SensorExcel users. Usage of the samples was high, at 95 per cent, as were measures of satisfaction (93 per cent) and recall (95 per cent).
Significantly, 78 per cent of sampled consumers said they would definitely buy or intended to buy the product in the future.
Nick Wells, managing director of CD, says that virtually all distributions are now targeted, but warns that complex GIS profiling may not always be appropriate. “I have one caveat. Often clients in the retail sector brief us to develop profiles of customers and penetration in catchment areas. Targeting like that is fine, but you have got to do it within the constraints of what else you want to achieve.
“There’s no point going down to one sector per store if you want mass-market coverage,” he says.
There is little doubt that GIS is beginning to play a major role in making local area marketing initiatives more efficient and profitable in the retail sector. Manu- facturers and retailers are increasingly working together to examine the profile of an individual store’s catchment area in relation to the profile of its customers. This can allow them to adjust ranging and marketing to improve penetration within target customer demographics.
“If you want to attract that different group, you have got to stock the products they want and then let them know about it,” says Roger Williams, marketing communications manager at Claritas, which carries out many GIS projects using its lifestyle data sets to identify product users and prospects.
He says that “the cost justification is from direct marketing. You use the GIS to target mailings and then see the increase in trade, sales and voucher redemptions”.
“We have found with a couple of organisations which have done this that stores get increased sales, store traffic and awareness, and the manufacturers are selling more of their products and taking a stronger position in that store”.
Some retailers are applying GIS to the segmentation of their whole chain into specific groups, rather than on an individual store basis. Mace used CACI to carry out a GIS profiling operation on its 800-store estate. Tracey Weir, director of the market planning group at CACI, says: “Stores were clustered into six categories based on the Acorn profile of their catchment area, how far they travel, and so on.”
These clusters were then passed to packaged goods suppliers to develop “planograms” to fit the profile of each group. One outcome was that some categories began to stock larger pack sizes, which Mace had previously assumed its customers would not want. “They saw quite significant upturns in sales,” says Weir.
Use of GIS may also result in an entirely new approach to business and marketing which can have a major impact on revenue streams. This was the experience of Cambridge Cable Group, which worked with Experian and its Micromarketing product to im-prove profitability, raise brand awareness, reduce churn and shape its product offer to meet customer demands.
With a major investment in laying cable, the company had done little to link construction with future customer uptake. With Experian’s help, it introduced a strategic marketing approach that established a financial link between profitable customers and the effective allocation of construction resources by segmenting its existing customer base and mapping this against the franchise area.
“In effect, there has been a step improvement in the quality of both internal and external communications which has brought bottom-line benefits as well as closer affinity to our customers,” says Sheila Munnelly, marketing product manager at Cambridge Cable. In a first step, the customer base was profiled using Mosaic overlaid with profitability and market penetration. This identified which were the most valuable target groups and where the areas of highest potential were.
“In addition, it established some facts about our customers that might not have been evident. Even though demand and initial take-up was good in the lower income groups, such as Mosaic type Low Rise Council, we realised that those customers were also most susceptible to attrition,” says Munnelly.
In addition to ensuring that cable was laid past the doors of the best prospects, product design was also influenced by the GIS exercise. By analysing revenue and churn within each market segment, it was recognised that customers leave for different reasons.
One of these was the need for extra service levels among the most profitable customer segment. High income customers do not have time for complicated programming, so a pre-programmed phone with simple to use functions was introduced.
Customer profiling, targeting and analysis can be carried out using existing systems which most marketers will have in place. “You can do it without GIS,” says Jonathan Plowden-Roberts, head of consultancy at The Database Group, “but it is a damn sight easier and more effective with it.”
It may also yield important insights which non-mapping systems could conceal. “You can’t do salesforce planning just by looking at a postcode. You have to see that BB is not next to BA – they are Blackburn and Bath, respectively,” he says. A simple postcode ranking might equally assign SW10 and SW11 to the same sales representative, even though they are divided by the Thames and might be more efficiently served by different people. In marketing, a picture really is worth a thousand words.