The Price of Marmite

There seems to be an entertaining public spat going on between two large global corporations about the price of Marmite. Unilever own the brand, and are trying to pass on the price increase to their customers. Tesco, who (reading between the lines) probably leaked the story, has refused. Or, at least, temporarily stopped selling Marmite on its website. The catalyst was the fall in the pound caused by BREXIT, but the roots of the conflict go deeper I think.
The different share price history of the two companies is revealing. Over the last 5 years, Tesco roughly halved in value from 400p to below 200p. This was caused by accounting irregularities, profit warnings, but the deeper cause seems to be competition from both German discounters (Aldi & Lidl) and the internet.

Even Buffett got Tesco wrong

Tesco should be able to defend itself from the competition. Normally you would expect a business like Tesco, with a high market share to have the lowest costs. That’s tough to compete against. This was probably why Buffett bought the shares. But even Buffett got this wrong. He overestimated Tesco’s “competitive moat.”  To his credit admitted his mistake and sold.
The conventional wisdom is that Tesco over expanded. The supermarket became too unfocused, trying to take on the banks, run a garden centre (Dobbies), a restaurant chain (Giraffe), not to mention the international businesses such as Turkey, South Korea and Thailand. This left them complacent and vulnerable to competitors who found a more effective way to serve customers.

A better explanation?

While that is true, I can’t help wondering if there is a better explanation. For a long time Tesco has been associated with loyalty cards program and “big data”. They bought the company, Dunnhumby, which is supposed to be an expert in customer loyalty. The DH slogan is the altogether cheesy “helping businesses love their customers.” The retailer claims that “Machine learning is ubiquitous, helping our lives become more enabled, more streamlined, more friction-free.” 
But it doesn’t feel like that to me. All this analysis of customer behaviour just makes our lives more complicated. Loyalty card points, vouchers that expire. Customers need a scientific calculator and a phD in gradient weighted moving finite elements to know if a UK supermarket promotion is worth it. Tesco are not the only company that do this; banks love to confuse their customers. Airline customer loyalty programs are a horrible disappointment. Easyjet invited me to join their customer loyalty program, but when I tried to contact them, they kept me on hold for 23 minutes before I gave up. It is like dealing with Soviet bureaucracy.  I hate them.
And you might wonder, if this big data stuff is all so clever, why is Tesco historic profit margin less than 2%? The company aspires to a profit margin of just 3.5%-4.0% by 2019/20? That is barely keeping their head above water.


On the other hand the German’s have a different approach. The German language has a word “günstig”, which translates as something like “good value/ favourable/ convenient.” Perhaps we should import that word into English. Günstig seems to be the German supermarkets expansion strategy. I haven’t spent much time in the German competition, but I’m told they have a narrower range and tend to push their instore brands. Customers associate these instore brands with good value/favourable/convenient. I also don’t see many people using loyalty cards in Germany.
While Tesco has languished, Unilever’s share price is up from around £20 to £36 over the same time period. And the company reports core operating margin of 15% at their most recent set of results, 7x higher than the supermarket and something that Tesco can only dream of. Dave Lewis, the Chief Executive of Tesco, just happens to be an ex Unilever guy. I’m guessing, but it looks like Tesco, know just how profitable branded goods can be, and want some of that margin for themselves.
I have no idea if it will work or not. But it seems to me that something much smarter is going on with Unilever’s brands.

Brands appeal to our gut instincts.

Brands appeal to our gut instincts.   We feel loyalty towards some brands. Brands work on the subconscious level.  Robert Cialdi, the social psychologist, quotes several academic papers showing how easy this is.*  Merely superimposing a Belgian beer 5 times with pictures of pleasant activities (sailing, waterskiing and cuddling) increased viewers positive associations towards the beer brand. Similarly superimposing mouthwash with beautiful scenes of nature 6 times led observers to feel more positive to the brand, both immediately and 3 weeks later.  In none of the studies were subjects conscious that they had been influenced by the associations.  If you want to see an ironic version of this, just click on this spoof of a generic branding video.  We might laugh (or cringe) but this stuff works!  No complicated machine learning algorithm here, but I repeat, this stuff works.

Customers identify with brands “marmite: you love it or hate it.”  Not everyone has to love your brand, but if even a minority do, there is good money to be made.

By contrast customers look at the “customer loyalty” programs created by Tesco, British Airways or Easyjet and asked themselves:

  • why make it so complicated?
  • why are they were doing this?
  • who it was all for?

Certainly not for the customer.  

Drowning in data

So perhaps these large corporations, that are collecting and analysing this data are missing something.  The Soviet Union had the best intelligence system, which could steal nuclear secrets, infiltrate rival agencies like MI6, and collect vast amounts of data on its own citizens.  Former East Germany’s stasi files are now available online.  But despite all this the top still didn’t know what was going on.  

In 1958 Nikita Krushchev visited a San Francisco supermarket and refused to believe the shelves were normally full, and had not been specially stocked for his benefit.  The fact that ordinary housewives could shop in abundance in the West, while Russian supermarkets has only queues, shortages and surly staff was a shock.  His intelligence reports probably played down the abundance of choice in the US. Gorbachev had to send in the KGB to Chernobyl, because the reports he was getting from the scientists at the scene of the nuclear accident consistently understated the seriousness.  The scientist were being entirely rational. In a society where anyone who contradicted the official dogma was sent to Siberia, of course bad news does not travel upwards.  

However much big data, intelligence and analysis you have, if you have a culture that is intolerant to diverse views, then your analysis is probably worse than useless.  I hope that the large amounts of data companies analyse can give them insight. Sadly my experience of mixing with bank management in the past, leaves me less hopeful about the value of customer analytics.
Interestingly, when Tesco tried to sell their big data company, no one was interested in buying Dunnhumby. So I’m sceptical that “big data/machine learning” really is going to serve customers. So far, if the airlines and supermarkets share prices are anything to go by, customer loyalty/ data analytics hasn’t even served investors.  As an investor and a customer, I’m sticking with the brands I trust.


*Academic papers quoted by Robert Cialdi in Presuasion :

  • Sweldens, van Osselear and Janiszewski (Belgian Beer)
  • Till and Priluck (Mouthwash) and
  • Winkielman, Berridge and Wilbarger (soft drinks)