Make sense of the data overload

Many marketers are failing to turn mountains of data into actionable insight that will drive the bottom line because they are not asking the right questions and do not have the right skills to analyse information correctly.

These damning observations were shared at the C-suite roundtable session sponsored by Microsoft and moderated by Marketing Week and Econsultancy at the Festival of Marketing.

John Norton, head of retail trading at The Royal British Legion, said: “In the years I have been in retail, I have found that we have now forgotten the ‘why’ – we focus on the ‘how’, ‘what’ and ‘when’ [but] I push for the ‘why’.”

Failure to ask the right questions is, in part, down to a lack of training and skills within many marketing departments, including those in the FTSE 100, claimed Katie Avon, former head of marketing and insights at publisher Pearson. On a year-long interim role, she found that analytical skills were lacking among her team of marketers.

“The challenge was expecting generalist marketers to try and do analytics.”

Katie Avon, former head of marketing and insights, Pearson

“When I arrived, I had a bunch of marketers who said: ‘Oh yes, we’re doing really good things with data and analytics’. But they were general marketers who had been trained up on Google Analytics. They could show me loads of presentations that told us the ‘what’ but there was nothing in there that was actionable or that was going to take the business forward. The challenge I found was expecting generalist marketers to try to do analytics.”

Clara Bermingham, head of digital at Rentokil, also believes marketers are not trained to deal with the ‘big data’ explosion. “There isn’t enough skill in the marketing department to be able to do decision analysis,” she added. That means marketing departments are sitting on data that could change the way the business operates.

Test your theories

Some businesses have circumvented this issue by forming partnerships with specialists who can provide actionable insight.

Working with the likes of Facebook, Mumsnet and Twitter, Mondelēz International is better able to understand “a bit more around the context of people” said Matthew Williams, former UK marketing director at Mondelēz.

But marketers need to ask the right questions to get information that can make a difference to the business. Williams’ former team had to change the way it worked in order to use information to drive campaign strategy.

He said: “We don’t have data specialists. And, after six weeks of getting data from partners and being overwhelmed with data, we switched it around and got our generalist marketers to come up with some hypotheses.”

One successful marketing campaign was born out of testing the hypothesis that parents treat themselves to chocolate once their children have gone to bed. Mondelēz’s strategy was to spend a small amount of money on testing the suggestion and when the analytics demonstrated the marketing messaging resonated, it invested more marketing budget into the campaign. This test and learn approach means the business can fail fast and move on, as well as scaling up quickly when the hypothesis is true.

Adam Johnson, CMO at Global Radio, suggested that once a theory has been tested, marketers can use it to develop predictive modelling. “[By] taking the data that your users are telling you and then modelling it around product development and commercialising it – that’s when it drives business growth and that’s what we’re all here to do,” he said.

One hypothesis tested for Classic FM is that people turn to classical music during key life events, such as births, deaths and marriages. Using IP data, the business can test that theory by analysing where people are tuning in to the radio station.

Johnson continued: “Then you can start to become predictive and look at birth rates, for example, and that period of time during the year when [classical music] might be the most relevant to them. That can inform our programming. And then there’s the commercial side. If you are Mumsnet, you might want to advertise around it.”

Attitudes behind the actions

The best actionable insight comes from combining different types of data, said Chris Capossela, global executive vice-president and CMO at Microsoft. “The magic happens where we connect behavioural data with attitudinal data and that’s where for our business it’s highly leveraged. If I know somebody is using Word and 20 hours of Xbox a day, fine, I get their behaviour. But knowing what their attitudes about the product are, what they like and don’t like [is key]. If I don’t have the attitudinal side, it’s a waste.”

Gary Booker, former CMO at Dixons Carphone, said that analysing data correctly is a “dark art”. The retailer has invested in econometric modelling and a small team of data analysts to enable it to understand what drives sales within the business.

However, he said it takes time to ensure you are analysing the data accurately. “The more you do it, the more you find what wasn’t quite right, the more you refine your parameters and the better you get at driving return on investment and knowing where you should spend it.”

Sometimes, measuring the data accurately is not about huge investment but about relationship building, said Emma Luetchford at Capita, who has recently moved from the marketing department to the IT team as head of digital transformation.

She added: “That’s where IT and marketing have to be best friends. My new tribe is the IT department – data scientists, very analytical and their skill set has to change because it’s not all about surveys and databases anymore. And then you have marketers who are very creative and, actually, the two can work well together but they’re such different tribes at the moment, so the challenge is bringing them together.”

Louise Herbert, head of marketing communications at Swift, agrees, and adds it is not just about sharing information and skills between marketing and IT. She said: “The trust culture is about ensuring that one department is not sitting on data. Collaboration is key – not just within organisations but with competitors and partners too because the more people that can have access to that data, the more intelligence and sense you can make from it.”

But Rob Stacey, head of marketing for computing at Dixons Carphone, believes that in the era of big data a new breed of marketer will need to emerge. He argued there is a new “skill set challenge” which businesses and marketing departments need to solve.

He said: “You need people who can handle the data but you also need people with the classic marketing skills in order to get actionable insight which will truly effect the bottom line.”