The success of the Tesco Clubcard in winning customer loyalty

International Journal of Retail & Distribution Management

ISSN: 0959-0552

Article publication date: 1 July 2004



(2004), "The success of the Tesco Clubcard in winning customer loyalty", International Journal of Retail & Distribution Management, Vol. 32 No. 7.



Emerald Group Publishing Limited

Copyright © 2004, Emerald Group Publishing Limited

The success of the Tesco Clubcard in winning customer loyalty

The success of the Tesco Clubcard in winning customer loyalty

An interview with Clive Humby by Sarah Powell, Editor of the “Spotlight” column in Emerald Now

Clive Humby is chairman of dunnhumby, a customer insight specialist whose client list includes Tesco, The Kroger Co. (USA), Barclaycard, Diageo, Gillette,, BMW, House of Fraser and Procter & Gamble. Clive works with clients on the strategic issues affecting their businesses while leading the company's work in the development of innovative media measurement techniques. He is also Visiting Professor of Integrated Marketing at Northwestern University, Chicago, IL.

Prior to founding dunnhumby, Clive Humby was chief executive of CACI Market Analysis. There he was responsible for creating and introducing a number of major innovations in marketing segmentation and retail marketing including the Acorn classification system and several major retail location systems. In total Clive Humby has over 25 years' experience in applying mathematical modelling methods to business, marketing and retail location. His book Scoring Points, co-authored with Terry Hunt, was published in October 2003.

Spotlight: What were the main features of Tesco's loyalty programme that enabled it to succeed when so many other loyalty schemes failed?

Clive Humby: Probably the main factor in Tesco's success has been the commitment to use the data gained to drive the business. Many companies have used loyalty schemes as a sort of quick fix, promotion tool. While, to a degree, a loyalty scheme is a form of promotion, it is an expensive way of doing this. The real benefit of a loyalty scheme is the very rich data obtained on customer behaviour. These data improve a retailer's ability to make the whole shopping experience more compelling, with the right products on the shelves at the right prices and with the right promotions. This use of data has been the major difference between the Tesco scheme and other initiatives that were less successful.

The volume of data we collect is enormous. Tesco's sales run into tens of millions of items a week. To make sense of all this means a total focus on using the data well. Many companies collect this sort of information but fail to use it in a committed way every day of every week. They spend tens of millions of pounds collecting the data but are not prepared to spend millions of pounds analysing the information gained. What they attempt to do is to derive the benefits from the sales and promotions effects of the loyalty programme, which will never work. Simply measuring how much the scheme costs and what sales are made on it does not tell you anything. If the sales achieved are not growing then the loyalty programme is not doing its job and there must be far more analysis.

Spotlight: What risks did Tesco take in the launch of the scheme and were there any features that were unsuccessful?

Clive Humby: Probably the biggest single risk taken was to go from the pilot of a few tens of stores to full national roll-out. Then there were the sheer logistics of making it happen – the challenges faced in acquiring and distributing sufficient plastic cards and training till clerks while keeping the details of the scheme under wraps.

One misjudgement was the original £10 pitch point for the scheme because many customers, and particularly retired people, made frequent visits, spending small amounts each time. But Tesco has always had what I would call a “test and learn” philosophy, trying out an idea in, say, ten stores and, if this proves successful, rolling it out to the rest. While risk can never be entirely eliminated, this approach mitigates the risk. It means Tesco is constantly learning, constantly gaining new knowledge as it goes forward.

Spotlight: To what extent and how rapidly did the card promote behavioural change among customers and how did this become evident?

Clive Humby: Prior to the launch, of course, we didn't know what individual customers did so we could not see any immediate behavioural change. However, a post-launch sales uplift in stores was an indication of incremental sales and thereafter we looked for consistency of behaviour, such as how long customers who had joined the scheme remained loyal to it.

While, on a week-by-week basis, there are inevitably shopping peaks and troughs, there are also underlying trends which can be recognized over time, such as family changes which lead to increasing and then decreasing baskets. For example, data analysis can reveal when children have left home, perhaps for university, because purchases of, say, pizza stop. Clubcard data offer many such pointers but the original measure of success is incremental sales, or “like-for-like” as we call it, i.e. how was the store doing a year ago? How is it doing now? Has there been a change since the launch of the scheme? Then the key measure over the years is the share of spend on the Clubcard, which indicates how well accepted the scheme is.

We cannot know for certain the percentage of customers who have a card, because we do not know exactly how many customers we have. What we do know is that 80% of total spending in our stores is made on the loyalty cards and this has been the case almost since day one.

Spotlight: What is the role of dunnhumby in the Clubcard scheme and in what way is Tesco's data analysis different from that of competitors?

Clive Humby: Our role as data analysts is to make the data tell the story that Tesco needs to hear, i.e. highlighting patterns in the data that give Tesco a commercial advantage over its competitors, such as buying trends and the results of promotions. Traditionally we could look at a promotion to see the sales uplift, but we would be unable to tell whether an uplift was the result of more people buying the same number of products, or the same people stocking up while the promotion ran, or new customers to the product category, completely new customers, or those switching from one brand to another. Understanding all of this is fundamental to working out which promotions work and then looking at other factors that we could build into the scheme, such as identifying brand-loyal people to see what can be done for them, or customers who like experimenting, others who focus on quality food, or those who have a tight budget, to see what can be done for all these categories of customers.

We can compare the Clubcard scheme to having a panel of ten million people whose buying habits we can analyse; as such it is invaluable to manufacturers. If, for example, a food manufacturer launches a new flavoured product, Clubcard data can show, for example, who buys the new product and whether they were already customers of the manufacturer's other products. The data also show repeat purchase rates, which indicate whether the new product can be considered a success. And because purchasing data are collected in real time from such a huge number of customers, they give manufacturers an immediate and greater insight into their products and range performances than could be offered through traditional, initial market research. The only question our data are unable to answer is “why?”. To answer this question we use panels of customers who have consented to take part in market research.

In recognition of the value of Clubcard data to product manufacturers and their different information requirements, we offer manufacturers and other interested organizations an independent commercial service that competes with traditional sources of market intelligence. We have, for example, conducted some interesting data analysis into the immediate impact on sales of product advertising on breakfast television or local radio. This sort of information cannot be collected through market research.

This arrangement means we are in effect data salesmen for Tesco and we also sell space in Tesco's media. This commercial arrangement helps subsidize the cost of the main process while aiding manufacturers to use Tesco media to best advantage and measure results.

Spotlight: How can Tesco determine the return on investment of its loyalty programme?

Clive Humby: We know how much money we give back to Clubcard customers, how much it costs to run the programme and what behaviour we get from those customers. We can estimate what behaviour we get from non-Clubcard customers through market research. Then we can look at the difference to work out whether customers with cards remain more loyal to Tesco and whether they redeem the offers we send them, and spend more when we give them money-off vouchers. We effectively come up with a series of conclusions which tell us what we can expect from the investment that Tesco makes in the programme.

Spotlight: What changes have been made to the scheme since launch?

Clive Humby: Apart from making the scheme simpler, introducing a straightforward £1 per point, we have developed a number of clubs within the club to respond to particular interests. While the general Clubcard is fine, for a young mother, for example, the baby club and targeted offers are far more engaging than a general communication from Tesco about everything in the store. Tesco has a number of such clubs within the club focusing on such areas as wine, organic and healthy foods as well as baby products.

All of these are “opt-in” clubs which customers can elect to join. They offer additional communications and are focused on a dedicated website. If, for example, you sign up to join the baby club, you will receive some £150-worth of vouchers for items that you would not normally buy from Tesco, such as prams, cots and so on, which are offered through the club via Tesco's dotcom business. This widens the appeal of Tesco, making its service particularly relevant to that customer's stage of life and interests. The birth of a first child marks a significant period of change in a customer's life and possibly the biggest change in family food shopping patterns until children are fully mature and “fly the nest”. So this is a fantastic point to engage with the customer and the retailer can make the engagement extremely relevant, supportive and focused through communicating baby-oriented information and special offers.

These clubs are part of the one big picture as Tesco simultaneously analyses data derived through the Clubcard and a customer's engagement with Tesco personal finance products. All this information enables Tesco to understand a customer's individual requirements and to respond to these with relevant offers.

Spotlight: To what extent has the loyalty programme aided Tesco's growth (i.e. to become the largest UK supermarket chain, the world's largest internet grocery supplier and one of Europe's fastest-growing financial services companies)?

Clive Humby: The Clubcard has been the technical tool to enable Tesco to build up a considerable insight into customer shopping behaviour. This insight includes knowing when and where consumers shop, which tells us something about how an individual or a family lives, and this in turn helps in terms of determining a launch strategy for financial service products. Most of these products are sold through in-store promotions with Clubcard point offers.

Meanwhile the Tesco brand is very strong and this image is reinforced because Tesco is perceived through Clubcard as being relevant, offering value for money and giving the impression of caring about customers. The points gained through the Clubcard are a compelling part of the proposition. In addition, Clubcard data can also help the customer wishing to buy online from Tesco. By giving a Clubcard number, items purchased on the customer's most recent visits to Tesco stores will be listed as a useful memory jogger.

Spotlight: What are the most challenging aspects of running the Tesco programme?

Clive Humby: The answer has to be the sheer scale of the operation. When we started out we were breaking the mould as regards what computers could do with the sheer volume of data we had – and you are talking about tens of millions of transactions every week. Even today, it is a huge challenge to make the data available quickly, usefully and at a price that's affordable, and to get great insights from this information. In the UK alone there are probably some 25 million Clubcard holders, which is equivalent to some 14 million homes. Approximately 10 million of these cards are probably active.

Spotlight: How does Tesco determine whether shoppers are “loyal” to Tesco or primarily “loyal” to Air Miles?

Clive Humby: That's easy. There are a substantial number of customers who joined the Clubcard scheme long before Tesco introduced Air Miles, which were previously offered by Sainsbury's. We know the people we attracted to Tesco when we won Air Miles from Sainsbury's. We know whether people convert their money offers to Air Miles or whether they use them for cash, spend them on treating themselves to holidays or other offers. We also know how involved customers are in the broader Tesco brand such as clubs and financial services. In addition we have a “share of shopping” model which predicts how much of their shopping basket we win. All of this added together gives us an excellent picture of general loyalty and loyalty to Tesco.

Spotlight: You note in Scoring Points that the customer relationship management (CRM) boom faltered in 2002. Why was this?

Clive Humby: I believe it is still faltering and I also believe CRM is built on a fallacy because customers don't want a relationship with their bank or their grocer or their supermarket. I've been in this game a long time and I remember back in the 1970s when store credit cards first came out and pundits forecast a “revolution” with customers using, for example, their Next fashion store credit card to buy all sorts of unrelated goods and services. Well, this didn't happen. The reality is that Next group is a fashion retailer and that's what customers want from it.

Tesco does not have a CRM programme. Tesco has a loyalty scheme and what this is saying is, “we get your data for giving you money back, and with the data we will give you a more relevant experience in our shops because you choose to shop there”.

Much of the hype surrounding CRM overlooked the customer. There was nothing in it for the customer. CRM became a means of doing things more cheaply – such as introducing a paperless relationship and offering the customer a £5 discount for signing up online. And many customers were underwhelmed by such propositions.

Customers will rebel when they see personal data being used against their personal interests, with offers made that are not relevant, or with offers made to new customers while longstanding customers are ignored. Where is the relationship in that? Customers will instantly recognize if they are being manipulated and this is something Tesco will not do. Tesco's aim is quite simply to do a great job for the customer.

(This interview was first published in Emerald Now,

Related articles