January 23, 2018

Lost in big data

Share Button
And the anwser is ...

And the anwser is … (Photo credit: jepoirrier)

British lingerie and swimwear retailer Bravissimo increased pay-per-click sales by 600% and doubled conversion rates over three months when it linked keyword bidding to weather data. The rationale for this action was Bravissimo found it sold more swimwear when it’s hot. This is just one example of the power of ‘big data’. Yet despite the potential for such lucrative returns, when it comes to the crunch, many marketers shy away from big data. It’s a phenomenon that seems, to many, so large and unwieldy that it’s hard to know where to begin, or end.

First, what is big data? In a useful overview, The Marketer states, “it is the search for commercially valuable insight hidden in repositories of data”. There is a huge amount of data available on a global scale, yet every piece can be pinpointed to a time, location, often even an individual. Pulling together masses of data in clever ways can indicate trends and patterns.

For example, this infographic uses weather data to illustrate global wind patterns, while this tracks real-time flights happening around the world. We’ve now reached the point where there are more ‘things’ connected to the internet than people who use them. From personal pedometers to radiation sensors, each ‘thing’ contributes streams of information. Thingful.net is a website launched at the end of last year that correlates information from all these devices, and displays data on a searchable map. So if you have an air quality sensor outside your house and you want to compare neighbouring streets, you can zoom in and search for similar sensors nearby.

As technology has become such a large part of our lives, our lives have become more trackable. Information from mobile phone apps, store cards, credit cards, social media, in-store tracking technology, can be combined and analysed for answers. Exploring big data can help organisations forecast sales, measure the effectiveness of online marketing campaigns, create better store layouts and merchandise displays, the possibilities are seemingly endless. Or are they?

The Marketer  highlights various challenges presented by big data. For a start, working with big data analysts is a new challenge for many. The trick is to be able to ask specific questions: be clear about what you want to find, maintain some focus. It comes down to that all important ‘aim’ of any research project. There are also numerous data protection regulations to take into account, which are neatly summarised in the article. Furthermore, controversy has arisen over the use of big data and finding correlation over causation. Data patterns might reveal that more burgers are sold in hot weather, but is this just a coincidence? And if not, why are more sold?

It is this last point that highlights the main limitation of big data for me. As with any quantitative analysis, it can only go so far; the crucial question why is not addressed. I think big data analysis is an essential tool in the researcher’s toolkit, and one that’s here to stay. But it’s not the golden arrow. Effective marketing requires a comprehensive approach to research that looks at the ‘what’ and ‘why’, and that means no amount of data can replace, in my view, a good old-fashioned conversation.


1866 Total Views 1 Views Today
Jane Priest, Teaching Fellow About Jane Priest, Teaching Fellow

Teaching Fellow at Edinburgh Business School. Part of the Marketing teaching team, focusing on the Consumer Behaviour elective.