Top

Tesco Shines at Loyalty: An Interview With Clive Humby 

Clive Humby | Mar 31, 2004 14,182 views No Comments

Share on LinkedIn

When dunnhumby began working with U.K. retail giant Tesco on what would become the successful Clubcard program, it was just a scheme of one Tesco executive to move beyond typical loyalty programs and tap into what the customer was actually buying—and would want to buy. Loyalty expert dunnhumby helped Tesco transform a wash of data into insight into the customer. Clive Humby, chairman of dunnhumby, speaks with CRMGuru founder Bob Thompson about Tesco’s loyalty program and the importance of creating strong relationships with customers.

 

This interview, conducted Jan. 28, 2004, was edited for clarity and length.

What is loyalty marketing?
Tesco’s Clubcard
Technology in play
Measuring success
A stepping stone

Bob Thompson
Thank you very much for spending a few minutes with me. I’m fascinated by the Tesco story. With all of the bad press that CRM (Customer Relationship Management) has received over the years, Tesco really stands out as one of the shining examples of how companies can really create better loyal relationships with their customers.

Clive Humby
Yeah, I think that’s true.

Bob Thompson
What is your job today, and how are you involved with Tesco?

Clive Humby
I’m chairman of dunnhumby. We’ve been working with Tesco since Clubcard was launched, and our role is to find the stories that occur in the very detailed transactional data that as a grocer, they get. And those are used, really, for two main areas. One is in the direct communications with customers. That’s about making those communications relevant to them—telling them about products they’re going to like or that they already buy or whatever.

The second is understanding how consumer behavior drives what I would call the retail decision model. In other words, which promotions work and why they work, not in terms of volume of sales but in terms of what they do for different groups of customers. So one promotion might, for example, appeal to cherry pickers who come into the store, graze it and leave it again. Another promotion might appeal to loyal customers. And, clearly, promotions that appeal to loyal customers do more about building long-term loyalty than cherry picking.

Bob Thompson
So the focus of your firm, it’s in this loyal marketing area?

Clive Humby
Yes, very much so. I think our job is to basically help our clients see their customers through their behaviors. And I think one of the bits of magic we bring is we convert their behaviors into attitudes through clever mathematical techniques that say, well, if you do this, this and this, then it probably means you’re on the Atkins’ Diet, for example, or it probably means—

Bob Thompson
You’re buying lots of beef?

Clive Humby
Well, exactly. And which customers are on the Atkins’ Diet, and more importantly, how long do they stay on it? And if they come off it, do they go back to the old diet? And understanding that’s very important in terms of projecting what’s going to happen to bread sales and pasta sales in three months’ time.

Bob Thompson
Can you give me a top-level view of what’s it like if you want to be in the supermarket business in the U.K.?

Clive Humby
Really there are five big players, about to become four. When we started Tesco, we were No. 2 to Sainsbury. Then there’s ASDA, which is a subsidiary of Wal-Mart, and Safeway and Morisson’s, And then there’s a couple of niche players that are primarily markets, Marks & Spencer’s and Waitrose, serving some very high quality to a few exclusive suburbs.

And there’s obviously the traditional discounters, the Lidls and the Aldis from Europe, but the vast majority of the market is made up of the big five, and shortly, Safeway, which is not related to the Americans’, by the way, is to be taken over by Morisson’s, which will basically be the big four.

There have been three players who have tried to use loyalty cards, Tesco, Sainsbury’s and Safeway. Safeway abandoned their scheme; they said they couldn’t mine the data. It was like having a fire hydrant of water when you wanted a glass. And Sainsbury’s gave up trying to run their own scheme and have gone into a cooperative scheme called Nectar, which is, as it were, a branded scheme that they, plus a number of other parties, participate in, Tesco being the only people who’ve successfully kept a standalone loyalty card working in their business. And I think it’s because they’ve used the data well, and it makes economic sense for them.

What is loyalty marketing?

Bob Thompson
Please define “loyalty marketing.”

Clive Humby
Loyalty marketing, to me, is about doing things for individual groups of customers that grows their lifetime loyalty. So recognizing that individual customers have specific needs, and that doing something for them and their peers is good for them, and it’s not the same thing as doing something for someone else.

Bob Thompson
So it’s not the same as just having a loyalty card or some kind of a points or rewards program?

Clive Humby
No. We’ve always said that anyone can launch a points system, but a points system is just electronic sales promotion. It’s just a different way of getting money off. It only becomes a loyalty mechanic when you use the data to add relevance to the customers’ life. And that’s what I would call loyal marketing, adding relevance to the customers’ life.

Bob Thompson
How is loyalty marketing different from, or related to, CRM?

Clive Humby
I think CRM is built on a foundation that customers want relationships, and somehow, knowing more about me means that I’ll suddenly want to do things with you. I don’t think that’s how customers work. Tesco has a phrase, which is, “It’s our job to earn our customers’ lifetime loyalty.”

When you start off with that mindset, it’s not what can I do that will make a quick buck from the customer. It’s what can I do that tells this customer I care about them. And because I care about them, they will stay with me for longer. I think the reason many CRM projects have a bad name is that they’ve been sold on the hype that knowing the customer was going to somehow magically make the customer more loyal. Well, it didn’t. In fact, what it tended to do was show up the flaws in the business because it actually made it easier to do things badly for customers. It [CRM] automated call centers because you were a lower-value customer, and you couldn’t get directly through to a person. Or it bombarded you with offers that you didn’t really want because it was cheap to send and you were clearly a good prospect.

Bob Thompson
CRM was more about marketing and prospecting than about true loyalty?

Clive Humby
Yes. To me, what loyalty means is, I’m going to do stuff for you and show you that I care about you, and because I care about you, you will trust my brand and you will respect my brand. And when I say to you, “I’ve got a good deal for you,” then you will believe me.

Bob Thompson
You’ve used the term, “I care about you,” a couple of times now. That’s more of an emotional component to the relationship, as opposed to, “I’m just giving you a good value,” or, “I’m giving you a targeted thing that’s very relevant.” There’s something in that “I care about you” that’s not so easy to quantify. How do you actually communicate that feeling?

Clive Humby
Tesco has their tag line “every little helps,” and that’s really the value we bring, which is: Everything you receive from us should feel like it’s a little bit of help. A lot of little things add up to trust, and I actually think it’s about the brand. I think, ultimately, it’s really about brand marketing, which is, “What does your brand stand for?” and how to reinforce what your brand stands for in the customers’ eyes in everything that you do.

And I think that’s very easy in above-the-line advertising and big-brand advertising. It’s much harder when you get down to the one-to-one communication with individual customers, where I might do something that’s perfectly sensible for most of my customers, but if 10 percent of my customers think it’s a dumb idea, then for them, I’ve got it wrong. And getting it wrong for that 10 percent is much worse than getting it right for the other 90 percent because the other 90 percent only care a little bit about it. But getting it wrong counts ten-fold when I’ll go, “Well, wait a minute. I don’t eat meat. Why are you sending me meat coupons?” That’s really getting it wrong. The converse is, “I eat meat. “You send me some meat coupons. Fantastic.” You don’t get that positive side, if that makes sense.

Bob Thompson
I want to refer to your new book. You’re co-author with Terry Hunt and Tim Phillips of Scoring Points: How Tesco Is Winning Customer Loyalty. In the front flap, it says that the Clubcard scheme, which is Tesco’s loyalty program, generates more than a £100 million of incremental sales each year, and that it has over 10 million members. And it’s profitable. This is a very big, complicated program. I imagine a sizeable investment in people, technology, etc.

Clive Humby
Yes.

Tesco’s Clubcard

Bob Thompson
How did Tesco create more loyal customers and profitable business with this Clubcard program?

Clive Humby
They started with a belief that understanding their customers was fundamental to running a successful retail business, and because they understand their customers, they hope they can serve them better. So what we’ve done, right from the beginning, is not look at this program as a marketing sales promotion program. But we’ve looked at the program as a way of learning about how consumers shop and what they want from their retailer and, therefore, not just changing the one-to-one communication with the customer but changing the physical retail offer.

Bob Thompson
So, you’re, first and foremost, looking for the insight into what really makes them tick?

Clive Humby
Absolutely.

Bob Thompson
And I take it from your earlier comments that you’re not just looking at this program as a clever way to discount.

Clive Humby
No.

Bob Thompson
So the insight was what really drove Tesco executives from the very beginning: “We want to know our customers so we can act on that.” Is that correct?

Clive Humby
Exactly. And I think it’s that board buy-in that’s been the difference. I think where many CRM programs have failed is they’ve been sold by the accountants as a “this is going to save us some money.” It doesn’t build the business around the customer. It doesn’t say is it good for the customer. If it’s good for the customer, it will be good for the business.

Bob Thompson
The story in this book is that the executive that was sponsoring this loyalty program eventually launched it in 1995. He had spent several years trying to convince the board to go in this direction, to make this investment. If it’s so obvious that customer insight is going to drive better business, why was it so hard for him to get Clubcard approved?

Clive Humby
Well, it wasn’t obvious until they found out. What happened was that they were used to looking at top-line numbers—”what are my sales this week? what are my sales last week?”—etc., and that’s the way they measure things. If you look at that level, you can’t see the impact.

Tesco brought us in to analyze the data, and we sat down in their boardroom—we obviously had done quite a bit of work before this—and we told a story, as we thought they would want to hear it. “Did you know that 5 percent of your customers accounts for 40 percent of this?”—which they didn’t know—and we then looked at some of the long-term issues. “Did you know that when you do this promotion, one customer in five comes into that department and starts buying and that’s new money into your business?”

It’s not about winning new customers, like in the credit-card market. It’s about winning a few more dollars in the wallet. There’s a lot of money to be gone after amongst your existing customers because they still go to other stores.

If you could understand that, and understand that intimately, that is how you are going to make your real gains. And I think we say it in the book. We presented that, and there was a long silence—and it really was long, and we were a very small company then, and, you know, we thought, “Oh dear, we’ve blown it,”—and then all of a sudden, there was a comment, “These people know more about the business in the few moments they’ve been working with the data than we know in 20 years of being retailers.” And that was it.

Bob Thompson
And were they clear on the fact that they knew it kind of by their experience and seat of the pants, and—

Clive Humby
Exactly.

Bob Thompson
They knew what was the right thing to do?

Clive Humby
They knew what was the right thing to do, and they’re a good retailer. The one thing I would say is: Any form of loyalty program can’t fix a business that’s broken. They already were a good retailer, and they’ve got all the basics right. And they were good retailers on gut instinct. But we showed them a way of getting underneath that gut instinct and saying, “This will be right in store A; this will be right in store B.” And that was the turning point for them, I think.

Bob Thompson
And this happened in 1995. What was the process of actually launching this program, and when did it really start to deliver value?

Clive Humby
We had been running a trial and started with 10 and grew to about 20 stores, and we built a business case which said if we could extrapolate the benefit that we’ve seen in those stores to the whole estate, what will it be worth? What are the risks? How wrong might we be? We build nearly all of our justifications on a sort of “this is what we think it will be worth, and this is the upside, and this is the downside type business case” now with the data. The biggest single challenge was finding enough plastic to print 18 million cards in total.

The second challenge for us was, obviously, scaling the technology we were using. I mean, we’re not a purely technology company. We’re really an analytical insights company, and the challenge for us became, “How do we scale the technology to manage, you know, 10 million customers, hundreds of millions of shopping baskets a year, billions of rows of item level data? How do we scale the business to do that?” And again, being analysts, we’re very pragmatic, so to begin with, we worked on a sample of data. We’ll find the patterns in a sample, and then look for that pattern amongst everybody, rather than just trying to find it in this huge data warehouse.

Bob Thompson
Do you think that may be, in looking back, one of the reasons this worked? That you didn’t have the resource to go build these gigantic data warehouses and then hope that data insight would just fall out of it?

Clive Humby
That’s right. We never asked for a check for more than $10 million. There was never a big bill without there being a big win already, and the biggest bill was, if you like, the leap of faith to do it. Everything else has been a little bill thereafter, and the leap of faith was made with a pretty solid business case: “This is what we’ve seen. This is the uplift we expect to get.” When it went national, and you had the benefit of national advertising behind it, we got even more of an uplift because, obviously, we couldn’t support a trial with anything national. It had to be supported purely with local media.

Bob Thompson
How long did this ramp-up period take?

Clive Humby
We were told at the end of November, and we launched in February. We had seven weeks.

Technology in play

Bob Thompson
I could see the problem with the plastic. Could you say a few words about the technology that you used?

Clive Humby
We started looking in the obvious places, like the Teradatas, but we couldn’t afford that. We have a hybrid of technologies. We use Oracle as our main data warehouse engine, to receive it and store it. We use SAS for a lot of our analytics. And we’ve used both White Cross and Sand Technology as our analytic engine for applying our learnings to larger volumes of data.

And on top of that, we’ve built quite a lot of our own technology to basically get the best out of the technology. So, for example, we could have built a massive single data warehouse. What we did, instead, was build a number of small ones that you could add the results of together.

So I could look at the 1-percent sample or 10-percent sample or 100-percent sample of the data, depending on what business issue I’m addressing. If it’s a top-line question, 1-percent sample is fine: Give me an answer in five minutes on an NT box. If I need to apply it to every customer and influence when we, for example, send the statement out to those 10 million customers, there’s now about 4 million variations of what we send. So virtually every letter is close to the one-to-one vision. We don’t actually do it on a one-to-one basis. We do it on a segment basis, using a number of different drivers. But to make that happen, we clearly have to apply it to the 100 percent data, and it takes a day on a big, analytic server.

Clive Humby
The analogy I often use, when you stand in the queue looking at the guy in front of you and what they’re buying in that supermarket trolley, you form a picture of them, don’t you? The clever thing that we do is see that in the trolley from the items that are there. So it’s not actually being able to track the items. The items, themselves, are irrelevant. We’ve got this algorithm that allows the items to describe the customer. That’s a clever thing, and that creates all these different languages that we use—these segmentations: the languages of the customer that say it’s convenience-driven; they’re concerned about their health; they’re looking for value for money; they like promotions.

Bob Thompson
So now that you know, through your data, how customers are going to behave, how do you act on that?

Clive Humby
Well, take promotions. There’s a lot of promotional money in food retailing. From a retailer’s point of view, you don’t really want to run too many promotions, because, actually, it’s very labor intensive, changing the shelf-end stickers, at a very simple level. It’s also very expensive in IT terms because, actually, promotions screw up your supply chain. You have a really big week on, you know, strawberry yogurt, and next week, you have too much because the automatic ordering system orders too many because you’ve just sold a lot of it, you know?

So what we’re very good at doing is saying, “Well, which promotions should I run? What do I think the impact will be?” And, “Am I seeing that impact, and, therefore, correct back the order for next week, so I don’t order too much?” Or, if I only want to run 100 promotions this week, which 100 give me best coverage of all my customers, so at least every one of my customers receives five or six things that will appeal to them? So how do I optimize—out of the thousands of promotions that the manufacturers would like to run—the right number to make it cost-effective for me in labor terms and great coverage, so most of my customers think that we run great promotions?

Bob Thompson
This is where your statisticians become handy?

Clive Humby
That’s right. So they’re building algorithms that say, “You want to do a promotion; you want to do promotion optimization; here’s the promotion optimization.” And we’ll build a promotion optimization that says, “Right. Put in your 50 promotions, Mr. Buyer, and it will give you the best selection of 12 for your department.” And we’ll predict how much money you’re going to make from that and how much money you should ask for from the manufacturer and all that sort of thing.

Bob Thompson
Then do you have tools that can drive these decisions into specific campaigns to generate email or direct mail pieces?

Clive Humby
Exactly. In a way, what retailers have become is a medium. If you think about it, they have their in-store posters and their in-store point of sale. They have their magazines. They have their loyalty program statements, and mailers and things. What we, basically, do is say that’s a communication channel to customers: “Hey, Mr. Procter & Gamble or Hey, Mr. Unilever or Hey, Mr. General Foods, which of those channels should you be in, and which customers should you be talking to about your products?” Because the philosophy we offer is, “Don’t try and switch a Coke drinker to drink Pepsi.” That isn’t in the customer’s interest. It might be in Coke’s interest, but it’s not in the retailer’s interest. And it’s not in the customer’s interest.

Find a way of helping the supplier see that if they engage with this group of customers, they can grow the category to their success, the retailer’s success. And the customer feels good about it, too.

Measuring success

Bob Thompson
What were a couple of really important metrics that you used to understand if the program was on track, if it was delivering the kinds of ROI that you were looking for or not?

Clive Humby
The biggest metric we use is the change in share of wallet. We have an estimate of what share of wallet we get from customers—in other words, what we think they’re spending. If I can see what you’re buying from me, I can sort of estimate what you’re buying elsewhere, if I know you’re a family of four or whatever it is. So one of the biggest metrics that we focus on is the change in the share of wallet. How many of our customers are effectively spending $5 more with us than they were a year ago? How many of our customers are still the same? How many of our customers have we introduced a new category to? And, basically, they see Tesco as being more than, for example, just food. Have we got them into clothing and other lines that we sell? So that’s probably one of the big metrics.

The other big metric is what we call our loyalty measure. But really, it’s a frequency measure, which is: How are we changing the frequency of visit of customers, and what sorts of visits are we getting from them?

Bob Thompson
So you can measure that directly?

Clive Humby
Yes, exactly.

Bob Thompson
The share of wallet is kind of more implied because you don’t literally know how much they’re spending with others. But you can tell if they’re changing their patterns, right?

Clive Humby
Yes.

Bob Thompson
At least by segment.

Clive Humby
Exactly.

Bob Thompson
We’ve found in our research with CRM projects that one of the critical areas is the role of culture, leadership and, in general, getting the people on board with the strategy. In 1995, when Tesco set off on this Clubcard adventure, what was the role of top management? How do they get people on board? Was that important, in fact, to the success of Tesco’s Clubcard?

Clive Humby
Exactly. I think you can’t underestimate the importance of internal marketing of the program. The top management at Tesco have always been passionate about it, and we knew we’d got it right when a couple of years after they started, their customer philosophy became … spelled out on every notice board in the company, “Our job is to earn the lifetime loyalty of our customers. It’s not being a grocer, our job is to earn the lifetime loyalty of our customers.”

That was the vision. And the other thing that happened is, when we launched the program, we actually sent out to the entire staff base a video which featured the marketing director and myself and some store managers describing the trial and what we’d learned from the trial. And that was shown to everyone, right down to every clerk on every till, and—

Bob Thompson
Everyone?

Clive Humby
Everyone.

Bob Thompson
That’s a lot of people.

Clive Humby
Yeah, nearly 200,000 people.

Bob Thompson
So they could have been bagging groceries, putting stuff on shelves, checking out, and, instead, they were watching a video”

Clive Humby
They were watching a video. It was only about eight minutes.

Bob Thompson
All right. Somebody probably took that eight minutes and multiplied it by all those people and said, “Oh my God”—

Clive Humby
Absolutely. But, what’s, I think, impressive about it was on launch day, supported by a marketing pack and the video, every single till clerk knew why we’d launched the scheme and what to do and what the script was and why it was so important to the company and all of those other things. And I think that is a very, very important part of the mix, which is if the staff aren’t behind the initiative, it won’t be successful.

Bob Thompson
How was it sustained over the subsequent years?

Clive Humby
Well, the store saw the benefit very quickly; the store saw the uplift in sales. I mean, literally, within six months of launching the scheme, Tesco had gone from No. 2 to No. 1 in the market.

Bob Thompson
So that certainly helps, doesn’t it?

Clive Humby
And that helps.

Bob Thompson
Show me some real business here.

Clive Humby
Yes.

Bob Thompson
And you will get behind it.

Clive Humby
Yes. Exactly. And everyone was really behind it. And I think the staff really were right behind the scheme from Day 1, and they saw it made sense. And I think a lot of them recognized it was not very manipulative. It doesn’t manipulate customers. It really is run for their benefit.

Bob Thompson
So it’s a more genuine or sincere interest in actually providing this value to the customers?

Clive Humby
They’re consumers, too. And they think, “Do I want this?” And it’s only if they want it, will it be successful.

A stepping stone

Bob Thompson
What role has this Clubcard program played in Tesco expanding into financial services, Internet shopping, maybe soft goods?

Clive Humby
Well, it was absolutely fundamental to launching Tesco Personal Finance because it allowed us to understand customers intimately, and what we started to do was overlay research on financial services’ behavior and say, “Well, can we predict financial services’ behavior from grocery behavior?” And surprise, surprise, you can. So that helped us develop and launch the bank.

I think, in terms of home shopping, it’s rather different. In a way, in the very early days, it didn’t help very much. It actually made things more difficult because you suddenly had to deliver this loyalty program through the Internet channel, which, you know, despite all the hype, is still quite complicated, sort of “it will work on this query engine, but it won’t work on that one” and all that sort of stuff.

But the real breakthrough came when we recognized that the biggest barrier to people starting being online customers is doing their first shop because you’re looking for all the things that you normally buy, and it’s a very long list—25,000, 30,000 items that a typical supermarket will sell. So, actually, we came up with this clever idea, which is: “Tell us who you are, and we’ll use your in-store list to populate favorites, and can you shop from that list. And that was a big, big win. Because, all of a sudden, we got double the click-through rates on first and second order because people found it easier to do.

Bob Thompson
Going back to one of your earliest comments in our talk today, you have other big, big competitors, well-funded, with lots of customers, lots of technology. How come they haven’t been able to do this?

Clive Humby
Lack of passion from the top, I think, probably is the biggest one. I think a lot of the people who run these big organizations see these programs as a cost, and it’s, constantly, “How do we do this for less?” That’s never been an issue for us. Certainly, if you get something you want to do regularly, it’s how you do it cost-effectively. That is a very fair question. But we’ve never had the issue, “Could we reduce the value by 5 percent or something?”

The second is a vision that the data isn’t just an electronic sales promotion. You have to physically re-engineer your business to be customer-insight led. I have to run my business in a new way to make it happen, and I don’t think many companies are brave enough to do that or have the skills to make that transition.

Bob Thompson
Yeah. It takes a bit of a leap of faith, doesn’t it?

Clive Humby
A bit of a leap of faith, but it’s like all these things. We have a test-and-learn philosophy. We’ve never asked Tesco to do anything big until we’ve proved it will work in a small way. So we’re constantly innovating, doing something in five stores, 10 stores; show it works; when it works, what does that mean? Where should we roll it out? And it’s that constant innovation; test and learn; and projecting where it will make a difference and where it won’t. And so many people, I think, are run on, “The average customer is worth $200.00 a year, what shall I do for them?” Well, that’s very, very wrong because if you have a very small number of customers worth $10,000 a year, getting it wrong for them can fundamentally alter the economics. And I think a lot of organizations don’t recognize that.

Bob Thompson
The key to success is leadership from the top plus attention to execution of these programs, whether it’s CRM loyalty marketing or most anything else that’s strategic?

Clive Humby
Yeah. I don’t think it matters what label you put on it, I think you can call it CRM; you can call it loyalty marketing; you can call it good old-fashioned marketing because that’s what it is.

Bob Thompson
Did Tesco say, “We’re going to do a CRM program?”

Clive Humby
No, no, the Clubcard scheme—Clubcard̬they view it as a brand. It’s a brand that means, “We care about our customers. We live our brand value of ‘Every little helps.'”

Bob Thompson
Yeah, so it was an integral part of their strategy to compete in the market?

Clive Humby
Yeah. And I think the other thing is they have never seen it as a technology issue. Many people think CRM means technology, and I don’t think it does. Technology is part of the analytics, possibly, and technology might be part of the delivery mechanisms. But actually, it’s culture, and it’s an approach. It’s a willingness to engage with the customer in that way. That is the fundamental issue.

Bob Thompson
Do you have some closing words of advice for our CRMGuru members, if they are thinking about loyalty marketing or a CRM program?

Clive Humby
One, make sure you get under the average customer. That is, understand the different groups of customers you have and think about those groups of customers.

Two, don’t think about how you’re going to sell your product. Think about what each of those groups of customers needs and have you got the right products for them? And if you haven’t, why not?

Third, if the management at the top don’t believe it, it won’t work. To me, if you get those three things, then, actually, it doesn’t matter what market you’re in, it will be successful.

Bob Thompson
It certainly sounds like having you involved with Tesco was a great asset to them, so congratulations and thanks for sharing your insights with us.

Print Friendly, PDF & Email


Recent Editor's Picks:


Categories: Chief Customer OfficerCustomer AnalyticsCustomer LoyaltyCustomer StrategyDigital MarketingEditor's PickEmployee EngagementInterview

14,182 views

No responses yet, why not leave yours?

Add Your Comment (All comments are reviewed by moderator, no spam permitted!)