In this new interview series, Starcount speaks to business leaders, entrepreneurs and pioneers across industries about their greatest successes, hard-learned lessons and visions for the future of data science and beyond.
Clive Humby is a data science innovator, with nearly 40 years’ experience in customer analytics across 30 markets globally. From creating the first geodemographic system, ACORN, in the 1980s, to transforming the landscape of customer loyalty with the Tesco Clubcard, Clive is acknowledged as one of the world’s pioneers in the discipline. With his business partner, Edwina Dunn, he founded global consumer insights businesses, dunnhumby, which revolutionised the use of transaction data for the FMCG industry, before expanding to work with a wide range of consumer-facing businesses, from banking and telecommunications to department stores and online retailers.
In the first of this two-part interview, Clive talks to Starcount about challenging orthodoxy, predicting future trends and the real value of influencer marketing.
After you sold your dunnhumby, yourself and Edwina Dunn took a long break from professional life. What drove you to come back to work?
Our first era was very exciting but we really only solved one problem – how to predict the future from a rear view mirror. And even then, it would only work for organisations with frequent touch points. Very few organisations can really understand what customers want and what they’re going to buy from historic transaction data. The grocers, credit card companies, mobile and on-line giants are the only ones who can – nearly everyone else gets a partial snapshot, at best.
I want to help organisations anticipate what customers will do next, buy next – to get more context about their customers and how to find likely prospects. I want to help them to understand the emerging trends, passions and motivations driving new behaviours, as well as who customers trust and what matters to them more than brands today.
What values and ideas have guided your career?
This new perspective challenges the orthodoxy I grew up with, which was that you take your transaction data and make it work harder. It has pushed me to apply new science and to use the latest AI tools and open source data.
There’s a very important distinction between doing the right thing for an individual and doing the right thing for a business. I worry about data getting too personal, but at the same time I think it’s vital for relevancy and efficiency that businesses have access to real-time understanding of emerging trends and needs.
You mentioned the idea of challenging orthodoxy. Would you say that’s something that characterises your career?
Yes, I think so. If you look back over my career, to start with, we were trying to classify people by where they lived. That was quite revolutionary at the time. Back then, we pushed the boundaries of what computers could do – and we still are!
None of us really want to receive messages from brands. However, we have to respect the right of businesses to promote themselves and we like free stuff – offers, apps, search. Without brands, we would be living in a grey world where we all buy the same thing from the same businesses; no diversity, little choice.
I’ve always believed passionately that if you can be really ‘on message’ to someone, generally speaking, the vast majority of people won’t complain, because they would much rather receive something relevant than be bombarded with messages and offers that distract and annoy. But, you need to give people control and that’s a balancing act that technology is now grappling with. Unfortunately, the law is probably two years behind technology capabilities.
There’s an online retailer that’s offering products with no branding or designed packaging whatsoever. Instead the products are labelled as what they are – olive oil or coffee, for example. How do you see the ‘brandless’ trend intersecting with traditional customer loyalty?
That’s a good example. Everyone would probably be quite happy with brandless products until someone revealed that the olive oil, for example, was farmed by low-paid, marginalised workers. A brand doesn’t just stand for the cost of a product or a logo on a shirt; it stands for a company putting itself behind the ethics of the product it’s selling. It’s a promise: “you can trust me”.
Of course, we’re in a global society and the ethics of communities around the world may differ and that’s where we get into an interesting dilemma.
Are you seeing many of those differences in terms of your new work outside of the UK?
The big themes are global but they vary massively from one country to another in how they manifest themselves. There are some things that are very specific to perhaps one or two cultures, but for the vast majority of themes they’re relevant everywhere. I don’t think there’s anyone who doesn’t care about the plastic pollution issue, for example.
On the subject of the plastic issue, Pret a Manger recently launched a scheme offering a discount to customers who bring in their own reusable coffee cup. The positive response from consumers triggered a wave where nearly every major high street coffee brand has jumped on board. How can you help to spot such a trend?
There are three big sources that can help you to spot when something major is changing. The first is that a particular business will be winning or losing in a major way. People talk about the things that matter to them, so listening or watching what is consumed on social is another way of spotting trends. And the third way is looking at how consumer behaviour is changing. However, that final source illustrates one of the challenges of the data revolution, which is that the systems are only as good as the data you put in them. If you’ve got no data on wrapping and packaging, for example, then you can’t detect a trend in it. That’s where social data comes into it.
Where do influencers fit into the picture?
The days of push advertising are nearly over. People now trust each to tell the truth about brands. I think the key questions are: what is the role of the influencer and who are they influencing? At the moment, influencer marketing is a bit like TV advertising was in the 1950s – all about the volume regardless of quality or relevance. However, where TV is effectively bought on demographics, I think influencers will be bought on passions.
Is it ever worth paying huge sums of money for celebrities with millions of followers?
There is no question that brands have always been successful because they’re associated with major celebrity. That’s been true since singers like Dean Martin and Frank Sinatra were around smoking particular brands of cigarettes and drinking a ‘favoured’ whiskey. I think we’re in a much more fragmented world now and brands have to ask what that particular celebrity represents for them. For example, I can’t answer whether Katy Perry with her hundreds of millions of followers is more valuable to a cosmetics brand than a beauty blogger with one hundred thousand followers; they’re performing two different roles. If you want to reach the broad twenty-something market, Katy Perry might be fantastic value for money. However, if you want to achieve an enduring purchase behaviour from a woman who’s switching from one brand of makeup to another, the beauty blogger will probably be more valuable in the long run, as they’ll have a more authentic connection with the target market.
If brands were limited to only knowing one type of insight about their customers, what would you say is the most important?
I think it’s about relevancy. Why would someone be interested in your product? If you can understand why they’re interested, you’ll be in a much better position to communicate with them.
Returning to the ethics around data usage, how do you see the fact that customers now have a better understanding of the impact of their own data changing brand-customer relationships in the future?
First of all, for most consumer marketplaces, the data for any one customer is not that valuable. GDPR is a very important initiative and I’m a big supporter of it – in principle. At the same time, it’s given people a false perspective of what their data is worth. If there are a million people who are in my marketplace and I’ve got data on half a million of them, the value of any one individual out of that half a million people that I know is very modest.
The exceptions to that are when you’re dealing with a very high value purchase like a house or a car, where data on the individual might be worth more, or when you’re in a marketplace where switching occurs like insurance or mobile phones, where your behaviour is a good predictor of profitability. Providers could use the ability to look at your mobile phone usage, for example, to give you a personalised price. In that situation you could see the consumer getting a real advantage from sharing their data.
Coming soon: part two of our interview with Clive Humby, where he predicts the challenges of the future, from technology to loyalty, and tells us his greatest professional headache and inspiration.