How an entrepreneurial couple helped start a retail revolution

GUEST POST from John Bessant

A gold ingot about the size of an older generation smartphone weighs 1kg, 2.2 lbs. To make it requires at least a million times that weight in ore, often a great deal more. That raw material doesn’t look particularly promising — it’s plain old boring rock, grey or brown with, if you’re lucky, some tiny tell-tale flecks of glistening yellow. But there is a lot of it about; gold comes from a variety of ores , often embedded in rocks like quartz which can be found anywhere on earth. Most of which is discarded in the extensive process of refining the metal, left in mountains of yellowing rock.

There may be great value in what finally comes out of all of this but getting to that shiny soft and heavy metal requires a lot of effort. The idea of lucky prospectors panning for gold and finding a pure clean nugget glinting away below the surface of the water is as far from reality as the presence of unicorns dipping their mythical heads to drink from the stream.

That doesn’t mean gold mining isn’t worth doing; that ingot is worth around $100,000 at today’s prices. But it does focus our attention on the importance or finding ways to mine and process the precious metal as effectively as possible. A kind of alchemy, transmuting base material into something of great value.

Which is what a couple of entrepreneurs started doing thirty years ago, developing tools and techniques for refining something similarly unprepossessing into a resource increasingly prized around the world. Knowledge.

Much like the raw ore which carries the high value of gold we have mountains of data available in various forms. The trick is to turn that unpromising source into high value knowledge of the kind which increasingly fuels economic growth and underpins effective decision-making in our organizations.

Edwina Dunn and Clive Humby were early prospectors. They met back in 1980 working at the London office of CACI, a company originally founded as the California Analysis Center, Inc. by a couple of RAND Corporation scientists who thought that using simulation and analytical techniques could prove commercially useful. Their instincts were good; from its origins in the 1960s the company had grown successfully and spread its work internationally.

Dunn and Humby were a good fit for such a knowledge-based organization; they worked on a variety of projects, drawing on his skills as a mathematician and her abilities in marketing. In particular she found herself working on the retail sector, trying to use data to help retail stores with location plans by developing improved understanding of their local customer bases. They were both successful; she rose from being a marketing assistant to become the youngest vice-president (at the age of 26) in the company with a team of 40 working for her.

But growing frustration led the couple to develop a plan to set up on their own; they formed a company taking its name from theirs — dunnhumby. The idea was that Clive would leave and Edwina would continue to support him from her senior position in CACI. A good plan in theory but one which soon crashed when unfortunate realities intervened. Within ten minutes of his submitting his resignation she was sacked by the company, leaving the couple sitting round a kitchen table in their home in Chiswick, west London, with some great ideas and a vision for how data science might change the world. But not a lot actually coming in to help them make ends meet.

Using her marketing skills and his knowledge of the key mathematical tools and how to use them they set about trying to promote their big idea. Their value proposition was around helping businesses unlock the hidden value in the data which they already collected and which could offer deep insights into their customer base. An early success came with signing the Booker cash and carry group as their first client, giving them a foothold in the retail world. But it also brought a major problem; their former employer began a lawsuit claiming that they were using ‘confidential knowledge’ which the couple had been party to when working for them.

In an interview Edwina Dunn explained that this ‘…..was rubbish, because we invented the knowledge, or certainly Clive had…”. But it put them under severe pressure; if they fought they risked losing their home and everything they had built up. “It was incredibly stressful. There was a big moment where we looked at each other and knew we could lose everything. They could break us. But we came to terms with the fact that if we lost everything, including the house and what we’d saved so far, we’d start again. That was a moment where you realize you’re quite strong — and you have to be in order to survive.”

Fortunately they eventually won their battle in the High Court and were able to concentrate on developing the business, bringing their vision of helping firms use data effectively to life.. One of the key advantages which they had was an understanding of how valuable data could be at a time when organizations didn’t. Indeed for many it was seen as a cost rather than an asset; once collected for whatever purpose it was too expensive to store, still less analyze, because computer power was still expensive.

dunnhumby’s alternative strategy was based on using data analytics to create a deep understanding of customers at a differentiated level so that clients could target sales promotions and engineer deals much more accurately. Over their first couple of years they worked from home, keeping costs as low as possible and paying themselves very little as they refined the plan and gradually grew the business.

A key challenge for them, analogous to gold mining, was getting access to suitable raw material. They needed datasets just as gold miners needed deposits of ore. Their big breakthrough came when they were invited to a meeting with the Tesco supermarket chain to talk about their ideas. Grant Harrison, a Tesco manager responsible for the rollout of a new loyalty card for the chain had seen Clive Humby at a conference and was interested in the ideas he was putting forward.

(Time for a quick detour into the wonderful world of customer loyalty programs)

Rewarding customers for their loyalty to a shop or a brand is not a new idea. Indeed it has been around at least since the eighteenth century when a US merchant began giving small copper tokens which could be redeemed for purchases at his store. In the mid-19th century the UK Co-operative Wholesale Society (the Co-op) began rewarding its customers with tokens which could be saved up and redeemed for cash or goods. The Great Atlantic and Pacific Tea Company began putting coupons in its packets of tea which could be redeemed for gifts in a catalogue; by 1915 customers could choose from over 60 luxury items on offer. The idea soon spread with an increasing number of retailers offering rewards for loyalty to shoppers in the form of tokens, stamps and points which could be collected. Frequent flyers were rewarded for traveling with the same airline, drivers could receive loyalty points to get discounts on fuel and shoppers could collect stamps to be redeemed for an ever increasing range of goods and services. New businesses emerged acting as the brokers, supplying the stamps or tokens and operating the schemes on behalf of major clients.

Data as a By-Product

The idea underpinning this long-standing business was essentially about getting close to and keeping customers; a by-product was the information that some of these schemes could reveal about customer identity and behavior. Something which the team at Tesco saw as a possibility when it was planning the launch of its ‘Clubcard’ as a points-based loyalty program in the early 1990s. Early trials of the idea suggested that in addition to the usual benefits of keeping customers loyal to the brand the Clubcard might also give them access to useful customer insights which could help future planning.

Harrison’s early work suggested that it might be possible to ‘mine’ the data about transactions made using the card but talks with major IT services companies suggested the costs and timescale would be significant. Faced with estimates suggesting a development time of three years and a cost running into tens of millions of pounds he thought it worth exploring what outsiders like dunnhumby might be able to offer.

Their initial offer proposed a 10-week development project costing around £250k and he decided to take the risk of seeing what they could come up with. dunnhumby developed a version of the Clubcard which was trialed in nine stores over a three month period; they presented their results at the end of that time and caught the attention of increasingly senior management. Eventually they were invited to present to the Tesco board; their report was met by a long and awkward silence.

It was finally broken by Lord MacLaurin, the chairman, who memorably captured the huge implications of what the couple had presented. “What scares me about this is that you know more about my customers after three months than I know after 30 years.”

What began as a short-term consulting project was transformative for both sides. dunnhumby’s work showed in detail patterns in what customers were buying, who they were broken down by various categories and identifying where further ‘data mining’ might be useful. Clubcard became the world’s first mass customization loyalty program in the world, offering a much finer degree of insight into particular groups of customers than anything that had previously been available.

That project became a long-term partnership from which both sides learned and were able to grow. Tesco’s success helped it overtake Sainsbury’s to become the top UK supermarket within a year of the launch of Clubcard. They not only benefited from their own use of the data analytics approach; in partnership with dunnhumby they signed similar deals with other supermarkets around the world. So successful was the Clubcard for Tesco that dunnhumby was soon approached to do the same for Kroger, the US chain competing with the giant Walmart.

In 2002 Tesco bought a 53% stake in the business and in 2010 bought the remainder. By that time the business was making profits of £46 million on a £248 million turnover. It employed 1,300 people across 30 offices worldwide and had other clients including Cadbury, Vodafone, Shell and Unilever. What had started as a kitchen table office and a real risk of bankruptcy for the two entrepreneurs had paid off to the tune of an estimated £93million.

At the heart of their original business was a simple belief — that buyer behavior wasn’t random but something which could be analyzed and the resulting understanding used to develop far more effective strategies for reaching and satisfying customer needs. With millions of customers the task of data mining was difficult but the rewards in terms of deep and tailored insights about segments and even individual purchasers would outweigh the costs in developing the necessary analytical technology. In a world increasingly driven by mass customization the potential for getting close to the individual customer and communicating with them, responding to their needs, anticipating their preferences and engaging their long-term loyalty offered a real strategic advantage to whoever could realize it.

These days we take the power of such analytics for granted; the spectacular rise of many of the big players on the global business scene like Google, Meta, Amazon and Alibaba owes a huge amount to such customer data science. Its tentacles reach beyond commercial transactions to the densely-populated world of social media and down into murky waters of electoral influencing and opinion manipulation. And with the rapid rise of machine learning the potential for deeply customized interactions individualized from a population of billions becomes a distinct possibility.

Dunnhumby offers an entrepreneurial success story demonstrating how a vision — in this case seeing the potential value in something others discounted or threw away — can become a reality. Dunn and Humby can claim to being pioneers in the world of data science and to have worked some alchemical magic, turning waste into gold. But it’s not a story of getting lucky; instead it reminds us of some key lessons about successful innovation management.

· First it involves much more than a big idea; it’s recognizing and shaping opportunity from the context around that idea. And it’s about both vision — seeing what’s possible — and what Angela Duckworth calls ‘grit’. Being able to put in the hard work to bring the idea to life and coping with the setbacks and unexpected challenges which the journey throws up. Perseverance and resilience are qualities which the couple clearly had in spades, not least when they were sitting round the kitchen table with no income, no clients and the threat of a lawsuit putting their self-belief to the test.

· Their success wasn’t built on a magic single idea which turned out to be just what Tesco needed — right idea, right place, right time. It was more along the lines of Pasteur’s famous dictum ‘ chance favors the prepared mind’. In their case they were able to respond quickly and effectively to the Tesco challenge because of the deep knowledge they’d already acquired developing and honing the tools of their trade.

Nor was it something which emerged overnight. It is a story of scaling a great idea through careful strategic development. At the core is a commitment to the knowledge base , the core competence which enabled them to enter and pioneer the field of customer data science (CDS). They hired smart people and built close relationships with universities who helped them identify the talent needed to contribute to the growing workload. When they started with Tesco they employed 30 staff and this number doubled each year over the next five. They created an academic partnerships program, developing research links with world leading institutions which has enabled them to stay abreast of the science shaping the future of their industry.

· Scaling innovation is a multi-player game and in many interviews the couple have repeatedly drawn attention to their commitment to developing partnerships as a way of growing. Their early and close relationship with Tesco was a deep and long-term relationship; one indicator was the level of trust which developed between them to the point where dunnhumby had access to all of Tesco’s cost information. They were able to see the profit margin of every product sold in the stores and with that kind of data it became possible to develop some of the elements of the Clubcard approach which gave it such a competitive edge.

They developed similar close links with other players like Kroger as they grew the business. Building a value network in this fashion enabled them to leverage resources, open up market access, and develop enhanced solutions and services. As Edwina Dunn explained in an interview looking back on how they successfully scaled their idea ‘….my best decision… was to do joint ventures with companies. Where they win, we win, and they make sure you never lose’.

Conclusion

There’s a famous old Yorkshire expression; ‘where there’s muck, there’s brass’ — which , roughly translated suggests that sometimes there’s real value in what might otherwise be seen as worthless. The challenge, whether we are talking about recovering precious metals from discarded scrap, re-using waste heat in sustainable energy schemes or seeing and exploiting the value in discarded data, is the same. How to turn what might be alchemy to robust and widely used science. A case well made by Edwina Dunn and Clive Humby thirty years ago and as relevant now as ever.

You can find my podcast here and my videos here

And if you’d like to learn with me take a look at my online courses here

And subscribe to my (free) newsletter here

Image credits: Dall-E via Chat-GPT

Sign up here to join 17,000+ leaders getting Human-Centered Change & Innovation Weekly delivered to their inbox every week.