Share this content
MyCustomer.com

CRM in Practice: dunnhumby

by
10th Dec 2006
Share this content

Marketing departments are among the most demanding devourers of data in any organisation. Effective and targeted analysis of data can drive bottom line revenue growth and provide a competitive edge over rivals in often fast-moving, high churn markets.

dunnhumby, the international marketing consultancy and services company, is tackling its data integration issues and looking to manage and store its huge volumes of customer data through use of technology from SAS Institute.

"What we do at dunnhamby is we explore a client’s data assets, identify gaps and suggest new business opportunities derived from the data," expains Dave Annis, UK data solutions director at dunnhumby. "We're a client-facing team, committed to providing innovative solutions. This is our business, and we can't afford not to do it well."

The consultancy had been using a home-grown solution, but as is so often the case in such scenaiors this had become increasingly costly to maintain and develop. With dunnhumby’s business rapidly increasing, a more robust, enterprise-wide solution was sought.

The sheer volumes of dunnhumby’s customer data presented a particular challenge. A typical client project for dunnhumby can involve working with tens of terabytes of data. "Our overall strategy is that we look after a lot of data and we know that the real value come our of joining up lots of data that wasn’t supposed to be going together," says Annis. "It’s getting insight out of it that is the trick. We’ve got a lot of experience in how to use customer data."

SAS’ solution has reduced the time taken to deliver these data solution projects from 6 - 3 weeks and decreased processing time from 24 - 8 hours. By switching to SAS, Annis says the firm is able to deliver a faster and more efficient service to its customers, helping them make better use of their customer data. improve loyalty and increase customer spend.

Using a recognised technology vendor such as SAS rather than its own home-baked alternative means that the opportunity to attract skilled staff from outside the company rather than having to train people up on how to use its bespoke systems.

"We have got data experts who use the system but when we were choosing the software we ran it with people who don’t have any particular SAS expertise and they were able to achieve results very quickly. Classic business intelligence tools in the classic sense are just a way to deploy data back to people. When companies just do BI they are missing a trick by simply deploying not much more than data."

SAS’ data integration solution has already been trialled on a retail project in South Africa with positive results. The first UK trial project began at the end of August and dunnhumby’s new joint venture with Casino is currently being implemented using SAS’ data integration solution. The aim is for the SAS’ software to be used in 10-20 projects over the next year.Dunnhumby already uses SAS analytics for projects such as the Tesco's loyalty card scheme for the British grocery chain.

Tesco, the UK’s largest retailer, boasts over 12 million customers for its Clubcard program, Data gathered such as name, address and other personal information as well as shopping and spending patterns has has helped boost Tesco's market share in groceries to 31%, according to market-research firm Taylor Nelson Sofres.

To help analyse the mountains of data that Clubcard generates, Tesco uses dunnhumby, which receives data on 15 million Tesco shopping baskets every week. Each product is scored on 50 dimensions such as price and the size of the package. The computer looks for customers whose shopping baskets have similar combinations of scores.

dunnhumby classifies shoppers in six segments. The "Finer Foods" segment, for example, is made up of affluent, time-strapped shoppers who buy upscale products. "Traditional" shoppers are homemakers with time to buy ingredients and cook a meal.

The data generated can be used by Tesco itself to judge whether new product lines are likely to be successful or even to decide on the location of new superstores. It's a prime example of effective business intelligence in action and delivering tangible results.

Clubcard coupon redemption is around the 40% mark and cost per redemption decreased. Since the Clubcard programme was introduced, sales have increased by 52% and outstrip the industry average. Store openings and expansions have increased Tesco’s floor space by 150% while tesco.com racks up over 500,000 transactions per week.

But crucially Tesco as an end user sees the value of sophisticated BI techniques; other firms still do not and there needs to be a change of approach.

"For a lot of CIOs, their role has been about being the guardian of the corporate data," argues Annis. "But what they need is a champion for the better exploitation of corporate data. There should be those who focus on BI and those who focus on running the business better. For most companies there is a huge data asset that could be used better for customer uses.

"We understand the IT issues and how do deal with the complexities of data and systems. Equally we understand the exploitation of data that marketing need to get the most out of it. You know, there is almost a role for a department somewhere between marketing and IT, a department that understands data and what you can do with it."

dunnhumby is an international marketing consultancy and services company that helps retail brands and organisations engage more completely and profitably with their customers. With experience of working with Tesco on Clubcard for more than ten years, the firm now focuses on the retail, FMCG and consumer service organisations around the world. Its mission is to enable clients to turn "unprecedented customer behaviour insight into relevant actions that demonstrably increase business value". Services include data insight and segmentation, marketing communications, media measurement, innovation, promotion, FMCG ranging, service initiatives and organisational change.time from 24 to eight hours.

By Stuart Lauchlan
News & Analysis Editor
[email protected]

Replies (0)

Please login or register to join the discussion.

There are currently no replies, be the first to post a reply.