Getting relevant

With 1,250 stores across the country and a 30% market share double that of its nearest rival, Tesco dominates the UK supermarket sector. But it has had more than a little help in getting there.

Inside the company’s marketing department, one partner is singled out as instrumental in catapulting Tesco from ‘just successful’ to ‘stratospherically successful’: consumer data analysis specialist  Dunnhumby.

When Tesco first started working with Dunnhumby in the mid-1990s, the supermarket was neck and neck with Sainsbury’s. But the introduction of the Tesco ClubCard in 1995 started a flow of highly valuable data that has enabled Dunnhumby to deliver deep insight into customer buying patterns and enabled the company to successfully expand not just into new supermarket formats but increase the scope of its goods and services to cover home electronics, clothing, loans and credit cards, petrol and scores of other areas.

Tesco is just one company that understands the power of analytics (other Dunnhumby clients include US grocery retailer Kroger, France’s Groupe Casino, Air Miles Travel and BSkyB).

What is becoming increasingly apparent is the difference between those that appreciate how that added value is created and those that don’t.

“The days when you could compete purely in price are behind us,” says Dave Annis, data solutions director at Dunnhumby in the UK. He gives an illuminating example. In its analysis of purchases around Valentine's Day this year, Dunnhumby found that consumers are buying cards much earlier than they used to: large numbers – especially, upmarket customers – are searching out cards in the second half of January or early February. At the same time, Dunnhumby’s analysts also found an unexpected spike in travel insurance purchases in the build up to 14 February.

How can retailers exploit such nuggets of insight? Newsagents, stationers and supermarkets would be advised to re-consider when to give space to Valentine’s promotions – expensive ones, in particular – although that decision needs to be backed by analysis of the impact of removing other product lines from the limited shelf space.

In the case of the travel insurance spike, Dunnhumby thinks it has spotted a growing trend among young men who want to surprise their girlfriends with a romantic mini-break abroad. The opportunity for insurance companies to tap into that market, pehaps through joint mailings with travel companies, is obvious – as long as they realise it exists in the first place.

20% response rate

Establishing such value is hardly a trivial exercise: it requires a vast amount of customer and transaction data going back several years and drawn from multiple sources. It also requires a team which understands what they are looking for and how to get hold of it.



The 12 million purchases at Tesco each week generate five billion data items.

And Dunnhumby currently holds 10 terabytes of that kind of data online, and leaves 30 of its highly skilled analysts “free to roam” through it, says Annis.

The upshot of such profound depth of analysis, according to Dave Dillon, CEO of Kroger, is “relevance”: The products stocked, the layout of stores, the promotions and so on are made more relevant and appealing to

customers not because of managers’ gut-feelings and observations (as would have been the case historically), but because of the hard data. As such, he says, analysis “replaces intuition with facts”.

In Tesco’s case, its knowledge of customers’ buying patterns means it can be highly relevant in how it targets them: the company has six million variations in the offers it can make to different groups of customers at different times. If that seems extreme, just look at the response: Tesco gets a 20% response rate to its direct mailings; compared to the UK business average of 1%.

“We can put that all down to a lot of hard work and analysis,” says Annis.

Actionable data

But achieving that relevance is not without its challenges. Embarking on such a large-scale business intelligence exercise often requires a “leap of faith”, says Annis. “This is still a relatively new field, so the most important question is who within the business will champion it.” In the case of Tesco, the backing came right from the top from the outset.

Another challenge centres around the threat to individual power bases within the business posed by the extensive use of data analysis. “Knowledge is power and people use data to protect their pet projects or to excuse failures,” says Annis. So companies often find they come up against resistance to initiatives where on-going decision-making draws on a wider set of data.

Most CIOs, says Annis, see the management and protection of data as part of their role. But how many have a “data strategy”, a strategy where the focus is on the end result the business needs and which unifies the multiple departments around their data?

“CIOs have a vital role to play in maximising the return from their data assets,” he says. “They need to evangelise the value, to focus on exploiting that data to determine business action – not just take responsibility for collecting and protecting that data.”

Pete Swabey

Pete Swabey

Pete was Editor of Information Age and head of technology research for Vitesse Media plc from 2005 to 2013, before moving on to be Senior Editor and then Editorial Director at The Economist Intelligence...

Related Topics