Big data? Any sort of data would be good

March 2013

You may have seen the recent BBC2 Horizon documentary ‘The age of big data’ but, if no, then perhaps you’re anyway familiar with the term. No to both? Well ‘big data’ is, as the term suggests, about a whole lot of data, but it’s more than that because it becomes ‘big’ when you combine data sets together and they become bigger than the sum of parts. Big data isn’t just about marketing data; it can be any sort of data – e.g. astronomy, crime prevention, financial services etc. However, in a marketing sense and as an example, if a retailer took separate data sets on their customers from their purchase patterns online, their social media activity, their response to promotional emails, their offline purchasing (so in-store) and their store card spending, then analysed all this data together to look for correlations and trends, then we’re talking about big data. So in other words when you look at everything you know about your customer at once rather than in isolation, you’re dealing with big data because it’s the combination which reveals all sorts of things which the individual data sets could never do.

Of course big data is relatively new because it’s the digital revolution over the last two decades – but accelerated in the last 5 years or so – which has generated so much data. And also of course, that trend will only continue as we increasingly rely on different electronic devices and online activity as integral parts of running our lives, with younger consumers who have only ever known a digital world representing an ever-growing percentage of the population.

I find this all very exciting in a marketing sense, because it furthers the extent to which subjective decision making can be removed from marketing plans. However, if my enthusiasm for big data is currently being tempered, then it’s by the fact that, for some clients, we’re still barely out of the dark ages when it comes to a willingness to embrace any sort of data. As an example, two years ago I was working on a brand and marketing strategy project and was able to get hold of some fairly detailed consumer data for the industry my client operated in. It was high quality stuff and I had no reason to doubt it whatsoever, but the client did because “in our experience it doesn’t work that way” – the context being how consumers researched their sort of product category. So they rejected it and stuck with what they’d always done (a marketing plan based on personal opinion).

What’s more, I’ve worked with other clients who have their own customer data stored in their system but are unable to access it fully, let alone manipulate it and come up with meaningful outputs which would allow for far better decision making.

What’s the problem then? Well many of my clients are family owned and run businesses, and data can be something which conflicts with the personal views of the owners and they’d rather trust their own instincts. In addition, family owned and run businesses don’t always have the level of specialist business skills you’d find in larger and non-family businesses (for obvious reasons), and this is particularly true where second generations of the family are in charge (the first generation made it work through their own drive, passion and original damn good idea). Compare this to large PLCs where if you can’t back up your decision with some data then you won’t have a job for very long, and you’ll start to see where the differences are.

So right now I’m on a bit of a mission on all things data, and early results are looking promising having recently convinced a client to invest in system development to allow their historic customer data to be fully accessed and manipulated. What’s it revealed? Heaps of good and useful stuff, and not least customer retention levels from different market segments over the last 5 years.

Family businesses can’t always learn from big PLCs – in fact it can often be the other way around –but on all things data they most certainly can. So if you’re not currently making the most of available data or even not using it all, then make that your next objective. Then once you’ve done that, start thinking about big data because that’s the future and your big PLC competitors are already on the case.