[Skip Global Navigation]

Customers

Customers Home

Whitbread

Situation

Whitbread has moved from being one of the UK's major brewers to the UK's leading leisure company, with a portfolio of over 4,000 hotels, pubs and restaurants, 87,000 employees, annual revenues of $5 billion and profits of $567 million.

Challenge

The leisure market is both expanding and changing rapidly. To maintain a dominant position, Whitbread must be able to anticipate changes and recognize potential for growth. It gathers substantial amounts of market research data. But traditional analysis methods have failed to provide a sufficiently sophisticated level of insight into the business changes in the market.

In particular, Whitbread needs to be able to:

Solution

Through the widespread use of SPSS within the research department, researchers have been encouraged to develop models to facilitate more accurate prediction and support better decision making. They are able to combine data from multiple sources: surveys, geo-demographic, marketing, sales data and so on, then present these to decision makers in easily digestible formats.

Results

Big organizations must be able to see the big picture, especially when that picture is made up of a mosaic of different sources of information. Whitbread's portfolio of brands is a roll call of many of the UK's familiar names in leisure such as Marriott, Pizza Hut, David Lloyd, Costa, Café Rouge and TGI Friday's, as well as famous beer brands such as Boddingtons and Stella Artois. It is essential that Whitbread is able to see both the big picture, and in great clarity, all the small pictures from which it is composed.

Over many years, Whitbread had built up a dedicated research department to service the needs of its 13 internal clients. Whenever any research is commissioned, the market research team obtains raw data from its fieldwork agencies, from which it produces its own tables and reports. The team had even run some more advanced statistical analysis on a mainframe system, for example to perform regression, correlation or cluster analysis. But because of the effort involved, none of this could be done on a daily basis.

With SPSS for Windows, professor Martin Callingham, as Whitbread's head of market research, realized that the benefits of statistical modeling techniques could become an everyday occurrence if the program could sit on the desktop alongside all the other tools his people used. He introduced it to his team, taught them, as non-statisticians, how to apply some of the models, and encouraged them to play with their data.

That was five years ago. Now, SPSS is an integral and vital part of their research process. The first thing the researchers point out is that they are not "statisticians". But the ease with which they can apply statistical techniques for refining and simplifying data within SPSS means that they can apply a freshness and sharpness of focus to the vast amounts of data they handle.

We don't think of SPSS as being anything out of the ordinary. But it has affected the whole way we do analysis.


A precise understanding of the performance and placement of its beer brands in the market to guide an annual budget of $400 million on advertising and marketing of these brands

By applying cluster and factor analysis to a usage and awareness study on beer, it has been possible to transform an enormous number of tables of figures into a simple series of instantly comprehensible radar charts that are as simple to read as a clock face. The model is able to impute whether an attribute is positive or negative to the consumer, rather than make false assumptions. Presented as a series of PowerPoint charts, brand managers can see instantly how their brand is performing along three dimensions: brand standing, product attributes and brand personality.

Callingham explains: "Before this, we did not know, and nobody else knew either. Previously, we would only be looking at actuals. You would assume a low score was a bad thing. But now we can say, 'no the market is saying that is not important'. An attribute can be differently important in different markets. From this, you can dig down and look at what it is that attracts different people to different brands."

80 to 85 percent accuracy in predicting future turnover when acquiring new properties to convert to restaurants

Over 300,000 meals are consumed each week in Brewer's Fayre, one of Whitbread's biggest restaurant chains. Realizing that this brand had different appeal in different markets, a rolling program of research was commissioned to assist it in opening up new markets and in refocusing the brand to different sectors. SPSS cluster analysis was used on the initial phase to determine the three most significant "need states" of its customers: price/value, quality and family-friendly. Further research established other discriminatory factors. Around 80 specific improvements were identified, and a rolling program of refurbishment is now putting these into action.

A rolling program of 80 specific changes to one of its major restaurant chains in response to a better understanding of the market, both nationally and locally

Researchers have also applied SPSS to a combination of the research data they held from 13.5 million customers and the geo-demographic data available from public sources in the UK. Using cluster analysis on location data, the unit has been able to carry out spatial analysis, and map out in considerable detail the profile of the local market of each operating unit. A model has also been developed to predict to within 80 to 85 percent accuracy, the potential turnover of each new acquisition.

"This is typical of the way we use SPSS every day at Whitbread," concluded Callingham. "So much so we don't think of SPSS as being anything out of the ordinary. But it has affected the whole way we do analysis."