Dunnhumby – The Wizard from the Tesco’s Data Mountain
(estimated reading time: 9 minutes)
For a long time, almost eons, Tesco – the biggest British retailer – was considered the most sophisticated proponent of Big Data Analytics. Its hailed success in collecting customer data and then turning it into ever-increasing value showed the way for a number of followers all over the world (Mercadona, Carrefour, Esselunga and the like – one of them, Mercator, a no.1 retailer in Slovenia, proudly wore the title of local equivalent of “Tesco in Britain”).
Much of Tesco’s success was closely connected to Dunnhumby. Wait a minute, who? Isn’t that the name of a character from a Dickens novel? No. Then it must be a wizard?
And, a wizard it is – though not from Hogwarts.
Dunnhumby or, actually, dunnhumby, is the most ingenious of all the data wizards of its generation. The startup, established at the end of the eighties behind a kitchen table (so the legend goes), has over the years mastered the secret tricks of turning raw data first into insights and then into action that changes customer behaviour in a favourable way, bringing the retailer that special something – a competitive advantage. A wizard, turning the invisible and untouchable into pure gold.
Said Clive Humby, half of the original dunnhumby: “I think our job is to basically help our clients see their customers through their behaviour. And I think one of the bits of magic we bring is to convert that behaviour 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.”
See. The magic, the mountain of data, the wizards. And the scheme:
RAW DATA > INSIGHTS > ACTION > Real change in customer behaviour
Back in 1994 Tesco was lagging behind the number one giant Sainsbury’s and was looking for a turnaround. One of the ideas to test was introducing a loyalty card – which in the years before computer power exploded had been a concept reserved for airlines and their frequent flier programmes. In an article on bbc.com, Edwina Dunn (the other half of the dunnhumby team) described it thus:
“When Tesco asked its IT department how long, and how much to do the [nationwide] roll out, they said three years and £50m.
“We said £250,000 and 10 weeks.”
The legend continued from there. At the end of the presentation of the first findings, Tesco’s then chairman MacLaurin broke the silence and expressed in awe:
“What scares me about this is that you know more about my customers after three months than I know after 30 years.”
The relation between Tesco and dunnhumby became the envy of every competitor.
Just check their brilliant progress in the Timeline:
1995: Dunnhumby helped Tesco establish the world’s first supermarket loyalty card, called Clubcard.
1999: Tesco doubled its pre-Clubcard market share, overtook Sainsbury’s and became the largest retailer in the UK.
Unanimously recognizing dunnhumby’s data wizardry as being instrumental in Tesco’s success, praise came from all over the place:
2002: Noelle McElhatton (Direct Response): One of the greatest mysteries of the data world is how customer insights from Tesco Clubcard’s data mountain help make it the UK’s biggest (and the most profitable) grocery chain.
2004: Richard Turcsik (Supermarket Grocery): Dunhumby’s Clubcard launched in 1995… has helped Tesco become the United Kingdom’s No.1 retailer, as well as the world’s most successful internet supermarket, one of Europe’s fastest-growing financial services companies, and arguably one of the globe’s biggest exponents of customer relationship management.
2009: Money Watch: One of the major weapons Tesco employs in its United Kingdom home market is a loyalty card programme developed with data mining firm dunnhumby.
Then came the crisis, which seriously shook both customers’ wallets and the retail world in general. But some players remained intact. As the Guardian commented: “Tesco halted the exodus of shoppers to cheaper retailers, by doubling the points available to shoppers.”
The Independent (2010) further underlined it: “Britain’s biggest supermarket owes much of its success to a husband-and-wife team who started a data-mining business in their spare bedroom.”
No matter, crisis or not, all the glory goes to dunnhumby. The power of using Big Data, executing data mining, using data profiling on the huge database of millions of people – of course, when done by the right data wizards – is unstoppable. Right?
In fact, the wizardry of dunnhumby was so praised that it seemed virtually unstoppable.
One of the anecdotes tells about a wife complaining about condoms on the online list Tesco provided. They couldn’t be her husband’s because he didn’t use them, she was absolutely sure about the mistake. The Dunnhumby analysts looked into the data and discovered that the husband did actually use them – but not at home. They apologized for the “corrupted file” but the moral was there: what wives don’t know, dunnhumby understands!
THE FALL BEFORE THE WINTER CAME
But then came autumn 2014 with the leaves falling… and among them a snippet of news that Tesco had recorded a 1.3% fall. Headlines like “Downfall of Tesco’s Market Value” soon followed. Then events picked up pace; Tesco’s turnover clearly fell and it produced a loss in profit before tax. Another shock to come. Mr Clarke – Tesco’s long-time CEO – was shown the door in 2014 due to “shocking losses at the UK grocery business.”
In 2015 the results were alarming: stocks fell to an 11-year low, Tesco posted a $9.6 billion loss for the previous fiscal year. People losing jobs, 43 stores closed. And yet more bad news to follow.
Wait a minute – this doesn’t add up. The world’s smartest retailer, Tesco, in trouble? How come? Who in the whole wide world could be smarter, more sophisticated?
WHO WAS THERE TO OUTSMART THE WIZARD?
Well, maybe we really can’t talk about sophistication. The new wizards were more seen as pedlars – guys going from town to town – not knowing much or anything about the big data. Their magic was really nothing fanciful. Just very down-to-earth, back-to-basics, almost raw and savvy-in-execution stuff. But these new guys in town didn’t really care about not having anything fancy to take out of the hat. And also they didn’t show up like arrogant brats – something crunching and munching the past data could easily lead to.
But surely the genial data wizard would have a trick to stop such transparent pedlars?
if only a wizard had put on the shoes to step outside the lab and see how real people really shop. Perhaps ask one or two of the lot what they really feel about something that Tesco likes to call a loyalty scheme? Why they can hardly wait to go to a place where they easily find the products they don’t want to waste their time with?
Someone could have washed off the conceit and have gone out there (where the truth is) – but not the data wizard, because from the beginning he was so infatuated with his magical procedure of turning invisible into value. He didn’t want to get his hands dirty out on the streets.
And so the German contenders, Aldi and Lidl – sneered at as “just discounters” – yes, the pedlars, had a clear path. Exactly what they needed to prove their blitzkrieg retailing ability. While Tesco was panting and struggling, the German lean machines were pushing forward with double-digit growth.
Warren Buffet, the world’s most prized investor admitted:
“I made a mistake on Tesco. That was a huge mistake by me.”
TESCO GOES ON – BUT SOMETHING IS MISSING
In September 2018, Tesco’s current CEO, Dave Lewis, opened a new format of stores, called Jack’s. It openly recognizes the need to cut the assortment from 35,000 to 2,600 SKU’s, and clearly states they will take on Aldi and Lidl as the “lowest cost in the market place.”
But there is something hidden behind or missing in this story that Tesco would like us to see as “getting back to the source”, to do what Jack was doing 100 years ago.
There is no more place for the wizard, once the most admired son. Yes, dunnhumby is completely missing from the reports of new beginnings. Just check all the Guardians, Suns, BBCs and other media of the world.
Jack, yes. Tesco, yes. Dunnhumby, no. Disappeared.
The wizard has left the castle and gone its own way. Partly moving to the immense shadowy forests of illegal data gathering with Facebook and the like. And gone is the aura that once surrounded him and made him untouchable.
So what can be learned from this mysterious omission in Tesco’s case?
1. The hype
The hype is like the air being blown into a balloon. Up to a point the balloon expands, then it expands some more and then suddenly – too inflated, overblown – pops. It might happen with a (big) bang, or just as probably with a sudden disappearance.
2. The one-sidedness
Big Data with immense data-flow shouldn’t – even if executed properly – become the only touch of an organization to the outside reality.
Actually, Big Data should be complemented by other data sources that provide valuable insights.
In our Linkedin discussions, Mike Anthony of engageconsultants.com says: “Tesco’s reliance on its own data meant that it missed some of the macro trends that were happening outside their stores … A lesson to us all in never looking exclusively at one data source!”
Yes, the good old store visits and field observations will not bring you admiration and envious looks, but they might still work.
3. The danger of the maths trap
Data analysis sometimes gets you so overwhelmed that you overlook the most obvious thing: Past data is not the best indicator of future expectations.
Says Ted Sarandos of Netflix fame, who established the business model around its elaborate Big Data algorithms:
“You have to be very cautious not to get caught in the maths, because you’ll end up making the same thing over and over again … It’s 70% gut and 30% data… Most of it is informed hunches and intuition.”
4. The blindness
This is the blindness of the castle noblemen where the wizard was employed as a resident. We might guess that Tesco overlooked all the warning signs because it was so self-reliant swinging along in the comfort zone that the reputed total omniscience of dunnhumby provided.
And also, was Tesco’s selling of the data itself its strength or the first hint of weakness?
5. Some good news
In a way, this fallacy of total control over customer’s actions might be celebrated as good news. Especially when taking into account the more murky territories of – let’s put it straight – illegal data collection recently revealed during “the Facebook affair”. This raises the question: has Cambridge Analytica really been as instrumental in Donald Trump’s success as it was believed to be, or might this view just be an excuse for his opponents who have lost touch with their voters themselves?
6. Accepting the shadow
But let this “downfall” not be accepted as a call for a witch-hunt, a case against all Big Data wizards – though the snotty attitudes of its executioners somehow call for just that. Dunnhumby made a huge impact, the Clubcard was a groundbreaking success – up to a point.
Still – there are two types of retailers, those with loyalty cards and those who do just fine without. And the former let better do something clever with the data collected. The wizards like Dunnhumby’s might be worth a treasure – once they accept the shadow, oops, its own limits, and recognize that there is a world outside their lab.
7. Tricks or treat
It might happen to any wizard. Maybe during the showing of he has forgotten that the tricks might be just that – tricks – and not the goal, the treat itself.
— THE END —
The case is developed for our Omnibus Workshops where we not only discuss the themes but also try different strategies on model stores and learn how to overcome the traps described above. We absolutely encourage learning through play and also show the ways of blending data with creative retail decisions for future success. Kindly invited to check it out.
You might check also our 2020 edition of Swift Category Management for Physical Stores in Digital Times.