This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Find out more here

Wednesday, 25 November 2015

Has big data made market research redundant?

Is MR now surplus to requirements? Or is the belief that big data can answer every business question misguided? Alex Chruszcz, Justin Sampson, Morag Blazey and Colin Strong went head-to-head to debate the issues last night at the House of Commons.

Turkeys don’t vote for Christmas, and you’ll never convince a room full of market researchers that big data will make their role redundant - despite an informed and impassioned appeal to common sense by Alex Chruszcz, head of insight and pricing at Asda.

Chruszcz was one of four speakers invited along to The Debating Group event in Committee Room 12 of the House of Commons last night to argue the motion: “Big data has made market research redundant”.

Chruszcz spoke for the motion – not, he said, because he believes market research lacks value. On the contrary, Chruszcz explained that data analysts and market researchers sit “side-by-side within Asda, working on projects together”. His budget is split 50/50. “I know that we get better insight by combining big data and market research.”

“Chruszcz said he is fighting ‘a losing battle’ to convince senior executives that both sources of information provide equal value to the organisation”

But Chruszcz said he is fighting “a losing battle” to convince senior executives that both sources of information provide equal value to the organisation. “What they hear is that big data can drive much more profitable commercial decisions.”

Opposing the motion would be wishful thinking, said Chruszcz. “Big data is making market research more redundant by being more useful and powerful.”

Big data is more plentiful too. “On any given day, my team are processing a billion points of data,” said Chruszcz. By pulling together all the sales and customer data, supply chain information, and weather and complaints data, “you get the what, when, how and who”, he said.

We still need to know the answer to ‘why’ though, said Chruszcz. Without ‘why’, all the other data is useless. And yes, for now, market research provides the why – but Chruszcz’s argument is that it won’t be long before big data is able to answer that question too.

Dealing with complexity
That seems unlikely, according to Justin Sampson, chief executive of Barb – the TV ratings body – speaking against the motion. Big data, by itself, struggles with the social nature of life, and the context in which events take place. “Facebook can track all your interactions with different people, but it can’t assign a value to those interactions,” Sampson said.

“Blazey’s argument is that data are facts – things that happened; things that can’t be argued with. And yet facts don’t always reflect the reality of the situation”

Asking people remains the best way to understand these aspects of human existence. But, said Morag Blazey, CEO of Ebiquity, speaking for the motion, the basic act of asking someone something automatically alters their response.

Blazey’s argument is that data are facts – things that happened; things that can’t be argued with. And yet facts don’t always reflect the reality of the situation, said Jane Williams, an independent consultant, speaking from the committee room floor in opposition to the motion.

Williams gave the example of the “sloppy shopper” – a segment of consumers who will happily buy an alternative brand if their usual one isn’t available. From the data alone, how would a brand owner know whether this was a regular purchase, an alternative purchase, or the start of a new, long-term brand relationship?

Depths of understanding
It falls to market researchers to pull together data from various sources to answer the type of question that was posed above. Indeed, Colin Strong, managing director of GfK Technology UK – who was speaking against the motion on the panel – made the point that big data is but another tool in the market research repertoire.

“MR is a pretty broad church,” he said, pointing to the Wikipedia definition: “Market research is any organised effort to gather information about target markets or customers.”

Big data fits that description – and it’s not rocket science, Strong said. “We can do it, and we can do a better job than the data scientists,” he said - adding that data scientists just “don’t understand consumers”.

“There is no discussion; no argument,” Strong said. “Big data is a resource that we will want to use.” And rather than making market research redundant, it will instead “reinvigorate the market research industry”.

The motion was opposed.

Our verdict

There was a lot of future-facing discussion last night – of what market research and big data will or won’t do – for a debate that was meant to be grounded in the present.

“Big data has made market research redundant,” was the motion, and it was right that it was opposed. As Alex Chruszcz argued, big data hasn’t yet done away with the need for market research. Maybe it’s starting to, and maybe it will do one day. But as of now, big data has not made market research redundant. And there were enough points raised last night to make you question whether it ever will.

Indeed, Morag Blazey did not envisage a future where research would be made completely surplus to requirements, despite speaking in favour of the motion. “Market research will shrink and morph and end up as a bijou niche within data-driven companies,” she said.

“Maybe it’s starting to, and maybe it will do one day. But as of now, big data has not made market research redundant”

In some respects, you could argue that’s already happening. I recently visited Ticketmaster to find out more about their insight team (and wrote about it for Impact Magazine). Ticketmaster does a huge amount of work with transactional data – it has 11m individual customer records within its database – and out of 39 people in the insight team, only two are dedicated market researchers. But while few in number, it was made clear to me that these researchers make an integral contribution to the work of the division.

Within Ticketmaster, market research happily co-exists with big data. The same is true for a business like Tesco. As Justin Sampson pointed out, Tesco has become one of the great British business success stories of recent years thanks to its investment in data and insight through the Clubcard loyalty scheme, managed by Dunnhumby. Tesco loved Dunnhumby so much that it bought the business outright. And yet it still spends a sizeable sum on market research every year.

Then there are examples like Google – the archetypal data-driven company – which has endless streams of big, behavioural data available to it, yet still sees the need for a market insights team.

And finally there is Twitter, a company that should – in theory – know everything about their users’ wants needs and wishes simply by monitoring what people are saying about them on their own service. So would it surprise you to learn that Twitter is currently recruiting users to take part in an online research community?

The Debating Group debate was sponsored by the Market Research Society. The debate was chaired by Nadhim Zahawi MP, the former CEO of YouGov. A calendar of Debating Group events is online at


Follow us on
Follow us on Twitter

Readers' comments (10)

  • Interestingly, Cambiar's about-to-be-published third iteration of the Future of Research study suggests that the two hottest topics for corporate researchers today are (1) Big Data and (2) consumer emotions. This suggests that intelligently designed research that gets not only to the why but to the emotions behind the why will continue to have its place, Big Data notwithstanding.

    Unsuitable or offensive? Report this comment

  • All the Big Data in the universe can't lever open the human heart. Data scientists might be able to accurately anticipate the spread of a flu epidemic up and down the East Coast of the USA. But why should victims visit YOUR drug store, buy YOUR medications, take a rest cure in YOUR resort, renew their health insurance with YOUR company, start to improve their resistance to disease with YOUR vitamin supplements? The quality of the product and the brand still counts, and Big Data simple can't unpick the inner life of consumers the way opinion research can. I believe there will always be a human role in interpreting data.

    Unsuitable or offensive? Report this comment

  • Clearly, there is a role - and a need - for both! Indeed we recently helped a retail brand to fuse both together. See here for more details...

    Unsuitable or offensive? Report this comment

  • Several thoughts to add.
    1. If Big Data is the answer what is the question? The business questions and subsequent research objectives will vary sector by sector, and by the specific aspect of the business you are trying to understand e.g. Advertising effectiveness, sales promotion initiative, change to brand logo, satisfaction with the call centre, new product development etc etc.
    2. It is yet another tool to use and not the only one. Ref Colin Strong’s points.
    3. Big Data does not have a conscience. It does not champion the voice of the customer on its own. So, we also need to think about who we recruit to manage and interpret big data. Is an expert in SAS the only pre-requisite? Do we also need a department filled with ethical and unbiased professionals, people who understand marco trends to help interpret the big data? People who can push back stakeholders who might work in a siloed manner or short-term perspective. People who can workshop the insights to enable action in the organisation. People who can see the Big Picture from the Big Data as well as enabling tactical solutions resulting from specific real-time data analysis.

    Unsuitable or offensive? Report this comment

  • There's room for both, and yet algorithms are catching up with qualitative analysis - I know, because we researched extracting personality traits from text and social markers for 2 years before we launched a commercial product.

    How do you lever open the human heart? Determine what's really in it by the words and actions of a person who doesn't know they're being watched - that's how.

    That being said, such results are validated and good, but still thin - you can get a much broader range of "Why" answers from asking questions.

    Unsuitable or offensive? Report this comment

  • I completely agree with Simon.
    Big data is a gift to market research as it allows businesses to have robust behavioural data. Equally, the process of understanding the 'why' behind people do what they do, and, crucially, what it means for the business isn't going to be replaced by algorithms anytime soon.
    What is crucially needed is a transformation of market research teams into competitive strategic insights capabilities (Time to stop being turkeys!).
    My experience of leading such a transformation is that if anything, you need more people and more human processes rather than less, if you are to create some sense and value out of data. You need them to connect the dots, to make the insights operational, and to deploy them in the business.
    I recently wrote a post about that for the marketing society.
    Here's a link if you are interested.

    Unsuitable or offensive? Report this comment

  • re-posting the link above ...

    Unsuitable or offensive? Report this comment

  • This is a very interesting debate. My own view is that for too long market research has been associated with simply asking people questions, whether by interview, focus group or questionnaire. It seems that transaction data is rapidly becoming another source for the market research / strategic insights team. All the points about interpretation seem correct to me. Data requires context and insight to create action!

    Unsuitable or offensive? Report this comment

  • Exactly what Tim said, its just another data source albeit a huge one with lots of potential sources - the assumption about market researchers is that we only use data gained from questions in a survey, which is a false start point. If you speak to anyone who works on STM's for instance you will know that to make those models robust you need actual sales data, its not big data in the sense being discussed, but its not data gained from a survey either and has been happening for decades.

    Unsuitable or offensive? Report this comment

  • Thought provoking post, and I agree with Simon. Big data is a boon in that it is another way to develop map of behaviors, to see what is happening. The part that Big Data has a much harder time with delivering is the ground truth of why those behaviors are happening. And generally the why gives direction on what to do about the what. MR, especially qual, fills in that why piece - Big Data, without a lot of work to filter out the noise and false patterns, is the continuing evolution of business intelligence and MR may well become the human side of BI.

    Revelation recently wrote up a case study showing a good example of how this works in practice. You can read about here:

    Unsuitable or offensive? Report this comment

Have your say

Please add your comment. You can include links, but HTML is not permitted.
Your email address will not be displayed on the site. All comments are moderated.

Mandatory What are the fourth and seventh letters of the word: majority