Did Tesco fall for the data analytics GIGO trap?
- Summary:
- Tesco's insatiable appetite for data driven decisions is looking decidedly shaky. Did they ask the right questions? Right now it looks like that's a good part of what has contributed to its current demise.
Is Tesco’s fall from grace a typical tale of shambolic succession and enterprise lassitude as times turned tougher? Or is it a market signal that Big Data, predictive analytics, and customer insight aren’t the sustainable competitive weaponry they’re cracked up to be? The schadenfreude gang may be counting on the former; but datanauts who referenced Tesco to sell their bosses on analytic investments would be wise to consider the latter possibility. Or is it probability?
I find this conclusion to a walkthrough of Tesco's woes intriguing on many levels. But let's wind our own clock back a tad. In March, Stuart Lauchlan talked extensively about Tesco's direction of travel in a digital age, noting among other things:
Indeed Clubcard is at the heart of the Tesco digital thinking, a loyalty program that is the jewel in the crown for the firm’s customer engagement strategy. Having 17 million Clubcard members data at your disposal, does make the goal of personalisation and that oh-so-elusive single view of the customer all the more tangible.
Tesco plans to make Club Card a “common currency” across all its brands and services – from banking to its Giraffe coffee shops. The group is also launching digital Club Card later this year to tap into growing customer use of mobile devices such as smart phones.
Riffing an earlier Telegraph piece, Schrage argues:
But the harsh numbers suggest that all this data, all this analytics, all the assiduous segmentation, customization and promotion have done little for Tesco’s domestic competitiveness since Leahy’s celebrated departure. As the Telegraph story further observed, “…judging by correspondence from Telegraph readers and disillusioned shoppers, one of the reasons that consumers are turning to [discounters] Aldi and Lidl is that they feel they are simple and free of gimmicks. Shoppers are questioning whether loyalty cards, such as Clubcard, are more helpful to the supermarket than they are to the shopper.”
Loyalty illusions
I'm a huge fan of club/loyalty cards but only when they're used in a way that benefits me first. So for example while in the US, I got myself a Ralph's card and I like it. They're a grocery chain which offers many of the in store discounts that come with club card ownership including the inevitable multi-buy offers. Registration is optional so I can silently opt out of being bombarded with whatever marketing crap they wish to push my way. But - and here's the kicker - when I check out, I see the discounts applied like a slot machine dispensing winnings. It's a great visual trick that makes me feel good about my shopping experience. Safeway does something similar. Tesco? I get points - whatever those mean.Thinking about the Tesco points thing, I see that as little more than a digital recreation of Green Shield Stamps. Anyone old enough to remember when Tesco used those? But more to the point (sic) I wonder if Tesco fell into the GIGO trap.
Garbage In, Garbage Out (GIGO) will wreak havoc on any data analysis project and from what Lauchlan is saying but for which Schrage offers a more blunt explanation, I suspect Tesco has been playing with the wrong numbers and then developing the wrong campaigns. Think about it for a moment. If you are losing market share then what do you want to know? I reckon it falls into two relatively simple buckets:
- Why are customers leaving us?
- Why are otherwise loyal customers spending less with us?
There are some subtexts to the second question, such as 'What offers are working for those customers? etc' but you get my drift. From everything I have seen, Tesco isn't asking those questions. On campaigns I have seen, Tesco has done little if anything to personalise the digital experience or, for that matter, to understand what my shopping preferences look like. I've never seen anything to suggest it asks the first question.
Instead, and as Zoe Wood at The Guardian argued earlier in the year:
Espirito Santo's Rickin Thakrar said Tesco was guilty of using "baffling" offers to bamboozle shoppers who had to do a "multibuy landmine dance" to get the best deal on fruit and vegetables. To save money on a price per kilo basis, including promotions, a customer would need to buy onions singly, peppers in packs of three, pears singly, packs of plums, loose red chillis and carrots, a sack of potatoes, loose broccoli, a pack of oranges, and individual mushrooms, he said of Tesco's pricing maze.
Compare that to what I see at Ralphs where there are often two or three prices as follows:
- Full retail
- Discounted retail if you hold a club card
- Discounted if you multi-buy on certain goods such as wine and beer (and who doesn't want to multi-buy those goods?!)
Each of these prices is clearly labelled and I have no difficulty in understanding my choices. Back in the UK - and Europe - German owned Aldi and Lidl are quite literally cleaning up and again, I can see why. They appear to have just about the right balance between choice and price to make it worth my while shopping there, even though many of the brands are unfamiliar.
Human decisions
In short, and like many others, I am fundamentally price driven for commodity items with little interest in brand loyalty. And, like many others, I cherry pick supermarkets based upon what I believe will meet my needs in changing circumstances. So for example, I am more likely to think Sainsbury or Marks and Spencer if I plan on putting on a dinner party. My real choice will depend on whether I can find a farmer's market on the day in question but that again is another nuanced element of the Tesco data problem.
And if you think that's a bit too folksy - then check this graphic from Gavin Heaton who talks about myths around retail disruption:
All of which leads me back to an ongoing discussion with Holger Mueller, analyst with Constellation Research. As I have noted before, he opines that 'true analytics' end up in actions. Maybe so - but if you are measuring the wrong things in the first place then you'll likely end up with some very poor insights, let alone actions. And if you're not understanding the changing buying habits of the customer then you really are up shit's creek with a broken paddle.
Right now, that's where Tesco is languishing and although a fresh take on analytics might help, it seems to me that without a clear understanding of the basics, they cannot easily recover. That in turn suggests Schrage is asking the right question but that it needs far more context then he is assuming.
As a side note - we could talk extensively about a changing of the guard and how succession plans are often fraught with risk but that's for another day.
Bonus points: video of my discussion with Mueller on analytics: