"As a result, the CIU contributes more and more to the company's decision-making process, at business unit and indeed at corporate level." The CIU now reports directly to the Group Marketing Director, Alan McWalter.
Second, Marks & Spencer has one of the richest and most extensive customer databases available to any retailer in the world. There are more than three million active M&S charge card accounts, and on average ten million transactions are made per week. Steven Bond comments, "For years we were sitting on a goldmine that went unexploited because we were already successful. Now we appreciate the immense value of all this data."
Data from the charge card system, combined with external sources such as census, demographic and national panel data tells Marks & Spencer a great deal about its customers. "We have at least 80 explanatory variables for every household in the UK, rising to more than 300 for any customer holding a charge card," says Bond.
Information collected at the point of sale, whether via cash tills, through the Marks & Spencer Direct marketing channel or the growing number of Web site interactions tells Marks & Spencer what they actually buy. "Multi-channel retailing is a major element in what we are doing, because when you put all of these channels together they tell you a lot more about your customers than looking at each one separately. Of course, you must have analytical software that can cope with the huge volumes of data we keep online. We are talking multi-terabyte," says Bond.
By applying analytical techniques such as cluster and discriminant analysis and data mining Marks & Spencer has identified 11 core customer segments (which are further subdivided).
The analysis helps in a variety of ways, from corporate branding through to operational decision making within the business units. It enables the company to ensure that the products in a particular store are the ones customers want. Whereas in the past stores were stocked according to their square footage now they are increasingly supplied according to a detailed profile analysis of their customers.
The new system supports the Business Units' decision-making process and helps to drive conversion rates and basket value.
Detailed segmentation analysis is also driving much-expanded communications and sales promotions activity. "We have a much better idea of what kind of offers to put in front of different customers and when, and what tone of voice to use, based on their individual tastes, preferences and behaviour," says Bond. "If we want to entice a regular food hall shopper into the menswear department, we want to know if he (or just as likely she) is the 'Egyptian cotton and silk tie' purchaser or has a lifestyle that demands non-iron shirts, for example.
Marks & Spencer has embarked on a more radical restructuring and reinvigoration of strategic store development, both to deepen and broaden the brand's appeal and to create more attractive, and "easy-to-shop" environment. Two-thirds of the company's retail space had been transformed by the end of financial year 2001.
Traditional gravity analysis and spatial modelling, which fuses SAS and GIS functionality, have long supported decisions about when and where to refurbish, redesign or replace stores to attract higher volumes of sales. Footprint analysis enables Marks & Spencer to predict the catchment area and any sales volumes of any planned stores.
However, these types of analysis usually assume a static, residential customer base, whereas today's strategies often call for a different approach. The company has recently introduced new trading formats such as railway terminal outlets and food-only stores. "The resident population around our recently opened food store at Liverpool St. station is negligible. Therefore our modelling is geared to predicting sales based on the flow of people through the station rather than the expenditure of local households. We rely on the flexibility of SAS to respond to such new requirements quickly and effectively," says Bond.
Careful analysis of customer behaviour also facilitates more effective marketing activity. For example, the shopping activities of the 11 segments over the Christmas 2001 trading period were found to vary. By identifying who shops and when – for example older customers tend to shop early to avoid the crowds, while younger men leave things until the last minute – the company was able to align its marketing activity and product availability accordingly. The ability to do this was one of many actions undertaken by the company to help drive up the successful Christmas trading figures.
"Our brand promise of trustworthiness and quality standards remains. But careful behavioral analysis enables us to be more flexible at the extremes, which is making a big impact on our financial results," says Bond.
Turning around the fortunes of a major retailer takes time. But with the help of SAS software, Marks & Spencer has gone to the leading edge of customer analysis. "Customer insight provides the fuel that powers the company's decision making," concludes Bond.