Play your cards right

Ten years ago, Tesco quietly began trials of a scheme that was to revolutionise its fortunes and help propel it to supremacy in the UK supermarket sector. Its ground-breaking Clubcard would not only reward and generate loyalty in its customers but also provide unique insight into their shopping habits, all for no net cost. In this exclusive extract from Scoring Points by Clive Humby, Terry Hunt and Tim Phillips, the inside story behind the launch of Clubcard is told for the first time.

In November 1993, the press began to report strange goings-on in the UK's supermarkets. "Tesco has started a customer loyalty programme", reported the advertising magazine Campaign. "The company" has introduced electronic 'swipecards' in three outlets. Although Tesco is playing down the launch of a branded 'ClubCard' as 'a test', it has already had a 50 per cent take-up from customers in all stores applying to join the scheme. Sources say that Tesco is now planning a national rollout from next year."

Whoever the sources were, they weren't official. Tesco's press office was denying plans for a national rollout of a loyalty programme. But the tests in Dartford, Sidcup and Wisbech had produced some remarkable results that suggested Tesco was on to a winner.

The origins of the Clubcard trial dated back at least three years. Tesco's marketing director in the early 1990s was its now chief executive, Terry Leahy, who had once spent a year working for the Co-op, the original purveyors of retail loyalty rewards. In 1990, having risen from the shop floor to the board at Tesco, he had been interested in the news that the Co-op in Bury was experimenting with a loyalty scheme based on a plastic card with a magnetic stripe. The Co-op was still awarding its customers a dividend - but in return it was collecting information on their shopping habits. Leahy was intrigued. "If they can do it in one shop in Bury", he asked his marketing team, "why can't we?"

With basic retail loyalty schemes common in the United States, Tesco's marketing staff had crossed the Atlantic several times already to see whether the idea would translate. Each time it was felt the start-up costs would be too high, and the chances of making an acceptable return on the investment doubtful. But by 1993, three things had changed.

The first was that it was now technologically possible to process the volume of transactional data a loyalty scheme would generate.

The second was that Leahy and Tesco were not alone in being intrigued by the possibilities. The long-term market leader, Sainsbury's, and an energetic management team at Safeway were also looking to the US. Sainsbury's had started a two-store loyalty-promotion trial, and if it was beginning to see this as the way forward, Tesco did not want to be left behind. Safeway was also testing customer response to a prototype loyalty card.

The third factor was that in the early 1990s Tesco was not yet the market leader. Pressured from above by Sainsbury's, and threatened from below by price cutters such as Asda and Kwiksave, Tesco needed to protect its current market share by sustaining and developing the value of its customer base.

THE PILOT
At the beginning of 1993 a £1m research and development budget was earmarked for the project. The implications of introducing a scalable electronic loyalty programme were well recognised by the business, as was the potential expense. The group's Electronic Point of Sale (EPOS) till system would need to be modified with card readers. A robust data collection and management system would need to be built. Data storage, analysis and security were vital. Staff would have to be trained. And the impact on in-store customer service would have to be planned for and monitored. Tesco would also need to create a compelling reason for customers to join, and design a marketing campaign to launch and sustain the programme.

The senior marketing team - Leahy, his marketing operations director Tim Mason, and Carolyn Bradley, now chief operating officer of tesco.com - were enthused by the potential of some sort of loyalty programme. Mason selected a young New Zealander, Grant Harrison, who had recently been appointed manager in strategic promotions, to head up the day-to-day development of what was beginning to be known as Clubcard. He was selected for his self-confidence and pushiness as much as his knowledge, but it did not hurt that his MBA thesis had been on the then little-known subject of loyalty marketing.

Harrison's first conclusion was that Clubcard would fail unless it was designed to appeal to customers' "hearts as well as heads". He argued that Tesco couldn't just use loyalty as a one-dimensional sales promotion aimed at encouraging customers to spend more. That would just be a repeat of the blunt and decreasingly effective scheme that Green Shield Stamps had become for Tesco by the late 1970s. "In my mind it was about two things: behavioural change and attitudinal change", he said later. "Combined, those two things add up to genuine customer commitment. And that's the most valuable goal for a loyalty programme. As soon as we'd agreed that, we knew our scheme was going to be different. If you just set out to improve and measure behaviour, we did not believe the programme would take root and develop. It had to be behaviour plus commitment." One of Harrison's first acts, in May 1993, was to appoint Evans Hunt Scott, now EHS Brann, a direct marketing agency. (For the first six months of the Clubcard trial Harrison had administered the programme alone - Tesco was still reluctant to commit resources to this unproven concept.) And on 18 October 1993 some impromptu customer service desks appeared in the foyers of Tesco stores at Sidcup, Wisbech and Dartford, piled with application forms, pens, a custom-built postbox and a modest display of posters (two per store) encouraging shoppers to sign up for their Clubcards and save on their shopping. They had to give their names and addresses and tick a box indicating the size and ages of their family, and in return they were rewarded with a plastic card, six vouchers for free cups of tea or coffee, a letter from the store manager and the promise of good news to come.

This was the trial Leahy had been agitating for. Codenamed Omega, it was intended to demonstrate whether or not a card-based loyalty programme was worthy of the board's attention. Many thought it at best a distraction, which was why it was modest, even by the standards of marketing pilot programmes, and why, even in November 1994, a Tesco spokesperson was still talking about "the potential to operate the scheme on a nationwide basis". Clubcard's success was neither guaranteed, nor even confidently expected.

For almost the whole of 1994, the debate internally was: could this thing work for long enough or on a broad enough scale to justify serious investment? To find the answer, the Clubcard team at Tesco had to work out exactly what they were offering, and whether or not that was compelling enough to get the majority of store customers to sign up and become committed card-carrying members. Harrison's research had shown that customer data was the big prize, and he was certain that sending rewards to customers at home was the best way to keep them interested. Jane Lacey, the planner for the trials, had come up with the vital ingredient: that Clubcard had to be recognised by customers as a reward for their existing loyalty rather than increased loyalty, a decision that was to have a dramatic effect on the long-term success of the scheme.

Yet there were plenty of decisions still to be made. How big a dividend or cashback should Clubcard offer? (The word discount was suspected of sending the wrong message.) What was the least they had to give away to get the most response? Should there be a "minimum shop" - a basket value below which Clubcard loyalty points would not be awarded? How about those who only spent modestly? In the early days of the trial, different offers were tested in different stores, because the Clubcard team had no idea what would succeed.

OMEGA'S FIXED POINTS
Some principles of Clubcard, however, had already been established, based on research or just common sense. The most important was that Clubcard needed to capture the customer's name and address. Tesco had the technology to track every item purchased from whichever store, at what price and when.

The only piece of the jigsaw missing was, who bought it? The capability was there. In 1993 the newest EPOS tills had been upgraded so they could swipe credit and debit cards. There was no new checkout technology needed to extend that to read the magnetic stripe on a loyalty card, although the programmers at Siemens Nixdorf had to write a new code for the EPOS tills. As Harrison points out, this was also a new concept for Tesco IT staff. "We also set up a small internal IT team who worked hard to understand what on earth we were talking about", he recalls.

Direct marketer Terry Hunt of Evans Hunt Scott described the challenge to the Clubcard team in their first meeting in June 1993: "My mum shops at Tesco Pitsea twice a week. She spends around £20 each visit. That adds up to £2,000 every year. Yet Tesco doesn't know who she is or where she lives or what she likes. In fact, it doesn't know anything about her at all. So what can you do if my mum decides to take her grocery 'account' elsewhere? Nothing." More worryingly, he pointed out, what can the manager of Tesco Pitsea do if a hundred customers like Hunt's mum decide to shift their "accounts" to one of the store's competitors?

The standard retail technique at the time was to send out untargeted promotions to everyone in the area, hoping customers who had deserted received them. But as Hunt says: "If you don't know your customers as individuals, it's like trying to play the piano in boxing gloves. You make a lot of noise but it's not always very pretty."

Without a name and address to specify who the member is and details of the household, Clubcard would be no more useful than a till roll, and no more motivating to customers than a discount card. With the member's name and address, however, Tesco was able to accumulate the rewards earned by their customers and send them directly to their homes in the form of money-off vouchers every three months.

TRIALLING ON THE CHEAP
This was only one of many marketing programmes that Tesco was trialling that year. At best, most supermarket promotional initiatives prove useful for a short time, or in a certain locality, or for a certain format of store. Finding one that can go national is the exception.

Harrison and his team didn't have a blank cheque. One of the attributes that Tesco had retained from its founder, Jack Cohen, was its thrift. Cohen called his autobiography Pile It High, Sell 'em Cheap for a good reason: he, and therefore Tesco, never spent more money than was strictly necessary. At the Cheshunt headquarters, marketing staff delegated to sort Clubcard applications manually had no money for office partitions, so they built their own from cardboard boxes.

Thrift was important in the economics of the programme - and that meant deciding how big the cashback percentage would be. A one per cent cashback would be cheaper but might not be compelling enough to change customer behaviour. Two per cent might be too expensive to pay for those changes.

Tesco wanted a scheme that appealed to the majority of customers, not just the highest spenders - partly because it wanted to be a democratic, inclusive brand, but also because of basic economics. If a membership programme, with all the investment that required, was to be worthwhile, it had to demonstrate an impact on the total business.

For many retail loyalty programmes, asking how many percentage points of reward would create more sales was as deep as they thought. They created a discount or dividend structure, rewarded shoppers for spending and trusted the goodwill they created would drive additional sales. Tesco's team thought this deferred discount was not smart enough. For its trials, the Clubcard team decided to go a step further by generating knowledge about its customers and using it to serve those customers better. They realised that with Clubcard, they would be able to look into its customers' trolleys, and that knowing what they bought and where they lived would provide a powerful insight. A reward scheme that does not analyse customer data might create a change in sales performance, but it leaves little scope to improve what the company offers.

By the middle of 1994, it was clear that customers responded well to Clubcard. In the the first week of the test, £6 out of every £10 spent in the three participating stores came from Clubcard holders. Over the next two months the proportion of sales attributable to Clubcard members rose to 80 per cent in some cases. Most importantly, customers responded just as well to one per cent of their spend as their reward as they did to two per cent.

For Leahy, this was a huge relief. If he was to make the Clubcard attractive to the board, he had to be able to minimise the risk to the business. There was no point generating higher sales through loyalty marketing if it resulted in an unsustainable dilution of margin: "Even with a one per cent discount, we would be giving a third of our retained profit away", he said. The marketing department may have been sold on the potential of Clubcard but, to convince the board, Leahy and Mason needed evidence that Clubcard delivered a lot more value to Tesco than an occasional sales blip.

That proof would come from customer insight based on the transactional data from the Clubcard trials. Tesco needed a specialist partner that could make its dream of data-driven marketing a reality. The Clubcard team knew that they would need to look outside the company's IT department to get the dedicated data analysis expertise they needed. So the search started for an outside company that knew how to take the massive volumes of transactional data and turn them into customer information.

In the early 1990s these weren't common skills. With Clubcard, Tesco wanted to respond to information on an individual customer's needs and preferences. To achieve this insight, it turned to another small agency of marketing data specialists: dunnhumby. The brief was simple: "Convince the board that Clubcard adds value to the business."

At the start of 1994, the only information Tesco could collect from the transactional logs was the date, location and amount each customer had spent. Tesco could see where the regular shoppers were, and who the big spenders were. But whether the big spend was on five bottles of champagne or 100 loaves of sliced bread was still a mystery. As the trials progressed through the year and were extended to 14 stores, the demand for more profound insight became more pressing. The Clubcard results would be presented to the autumn chairman's conference, to a group of directors who needed to be convinced that their £1m had been well spent - convinced enough to give the go-ahead for millions more to be spent on a national launch.

Luckily, dunnhumby's first look at the sample data revealed an overlooked but useful basis for analysis. The Clubcard-linked transactions at the till could be broken down by spend and department. Although Tesco was still some way away from being able to tell if its trial customers were buying Coke or Pepsi, the Clubcard data could reveal if they were buying soft drinks or nappies. Clive Humby, dunnhumby's "chief data doctor", and his team were very excited by the implications of cutting the data that way. Records of Clubcard transactions were delivered to dunnhumby's offices, and for three months over the summer of 1994 a dedicated team looked into what discoveries would most impress the Tesco sceptics.

They found a simple, yet at that time unthought-of statistic: that a small proportion of its customers accounted for a massive part of its profitability. These "premium loyals" would have a massive influence but before Clubcard they were anonymous; neither store managers nor marketing experts were aware of their economic significance. Other insights showed the profile of different customers based on the combination of store departments they used, highlighting sections that were failing to attract customers who shopped heavily in other areas. By analysing the penetration of Clubcard membership by postcode, the data gave a sharply defined picture, store by store, of how far valuable customers were willing to travel to shop. It showed in which neighbourhoods the competition was really biting, who was coming back several times a week and who saved up their shopping for a weekend.

Ten years ago even the most advanced retailers were unaccustomed to analysing customer behaviour. They were used to measuring footfall (how many people had come in and out), or the number of checkouts active during the day, or the number of baskets they processed, their size and value. This was different. This was what was really happening, not an average.

While the number crunchers were excited by the potential of Clubcard data, the bean counters at Tesco were a tougher audience. Tesco had made massive strides under chairman Sir Ian (now Lord) MacLaurin's leadership, but the fact remained that in 1994 it was still second to Sainsbury's. The argument for Clubcard was simple. By replicating the sales growth that Clubcard had created in its trial, Tesco could become the UK's number one.

On 22 November 1994, Tim Mason, Grant Harrison and Clive Humby presented the results of the Clubcard trials to the Tesco board. The board listened. Finally Sir Ian spoke from the chair: "What scares me about this", he said, "is that you know more about my customers in three months than I know in 30 years."

At the end of the meeting the decision was made to rollout Clubcard as soon as possible. On Monday 13 February, it was launched in every Tesco store in Britain.

THE NATION'S NUMBER ONE
Before Clubcard, Tesco was the UK's second-ranking supermarket. Today, it is its largest grocer, with an 18 per cent share of the market. Not only that, it sells more medicines and toiletries than Boots and Superdrug combined; more chart CDs than HMV and Woolworths; and 4.4 per cent of all UK clothing. In total, £12.30 of every £100 spent on the UK's high streets goes into Tesco's tills.

It is also the world's most successful Internet supermarket; one of Europe's fastest-growing financial services companies; and arguably one of the world's most successful exponents of customer relationship management. In the UK, its 221,000 staff make it the country's largest private employer, yet almost half of Tesco's floorspace is now outside the UK. And Clubcard is already used in stores in the Republic of Ireland and South Korea, with plans for others to launch it too.

No one would claim that Clubcard is exclusively responsible for Tesco's success. But talk to any of the senior executive team, and it is clear that the business benefits of Clubcard are now written through the business like lettering through a stick of rock. Clubcard for the first time allowed a mass retailer to know its customers personally and establish a long-term relationship with them. Four times a year Tesco sends its 14 million Clubcard-holders "money" - vouchers they can spend in the store. In the eight years between the launch and the end of 2002, the value of those vouchers totalled more than £1bn - yet Clubcard pays for itself, and Tesco makes a profit out of doing it. Since 1995 it has covered the cost of its loyalty programme with a sales uplift directly attributable to the promotions created by Clubcard. In short, Tesco runs Clubcard, and has been doing for seven years, for no net cost.