A survey by the Economist on the usage of data within companies found that 1 in 4 admitted that vast quantities of useful data went untapped, with 1 in 2 conceding that they leveraged less than half of the data they had available. A successful big data strategy requires organisations to overcome the challenges of Volume, Variety and Velocity of information to arrive at better decisions. Data fact or fiction? Web 2.0 has certainly added to all of the V’s in this equation, but how do we separate fact from fiction? And how much unnecessary cost has this added layer of complexity added to your bottom line? In a psychology experiment carried out by Leeds University Business School experimenters asked people what snack they would like to have in a week’s time – a banana or a chocolate bar. Most people made a banana their advance choice. The following week the experimenters returned and offered the same people a snack to have straight away. No mention was made of their previous choice. Most people, especially women, opted for the chocolate. Had this not been an experiment, but instead, it was a social media campaign with my target audience all “liking” my Bananas, how much time and energy could I have spent analysing what turned out to be aspirational data? Worse still what if I had used this to make or influence product or production planning decisions? I’ve used figure 1 below to reclassify data into three new categories: 1. Facts. Based on verified information and transactional data. 2. Aspirational Data. This is based on what your customer shares publicly via social media or privately through market research and surveys. 3. Predictive Data. Using statistics and algorithms to predict customers buying patterns and behaviours. Figure 1 1. Facts should be the #1 priority in any data strategy The Economist’s research also showed the 1 in 3 organisation still don’t have a formal process around data management, with a further 27% investing time validating and scrubbing their data. 34% of all respondents rated data quality as very problematic. So, before we rush into adding to our volume, and variety with web 2.0 and social media, there is still a lot of ground work and competitive advantage to get gained by getting the basics right. I attended a seminar a number of years ago where the speaker was the head of strategy for a global organisation. He was very proud of the fact that his organisation had ditched the traditional customer segmentation model for an event-based model. This relied heavily on a solid foundation of transactional and factual data allowing him to market to his customers only when they were in the buying cycle. (This had a massive impact on both the efficiency and effectiveness of all of his marketing activities) This could be the birth of a new child (based on nappies suddenly appearing in my loyalty card data) or an upsell to a “Gold” or “Platinum” credit card based on an identified increase in my salary from my banking data. A solid foundation here is also the key to unlocking the potential and value of predictive data. (see 3) 2. Aspirational data big picture only If you look at my Pinterest boards (and loads of other people boards). They are full of aspirational data. My “dream garage board” would cost me millions to buy and there’s not one car on there that I actually have the cash to buy. It’s the same story on Facebook and Google +1. In the same way if you asked me what I wanted for a snack I would probably say a banana (aspirational data), but when it came to making the choice I’d probably choose the chocolate bar (factual and predictive). From my perspective aspirational data can be hugely valuable in identifying big picture trends, and as a cost-effective way of validating new products and concepts (see my post on Lean Start Ups) Would I place a high level of reliability on it? And invest heavily in analytics and integration? Probably not. 3. Predictive data a source of competitive advantage Having not worked in an intelligence agency, been a criminal profiler, nor do I possess savant abilities in maths, means this is the area I’m least qualified to talk about, but believe has the most potential. Pictured above is my New For You list from Amazon. I’m not sure what this says about me, however it is based on factual transactional data (not allowing for gifts I’ve bought for friends and family – that’s my excuse and I sticking to it!) and it generally comes up with some very relevant recommendations, which result in me buying more books than I would have done otherwise. In figure 1 above I’ve placed a pot of gold at the intersection of Factual, Aspirational and Predictive Data, as this is, in my option is as close to perfect information as we can get. Now if Toyota can apply my model to use my Pinterest (aspirational) profile to establish my love for sports cars, filter out based on my factual data that I don’t have the means to buy a Lamborghini Aventador, but instead offer me a great deal on the more reasonably price GT86 then I would be a happy man.
Chocolate or Bananas? Which one will you invest in when it comes to your data strategy?
July 18, 2012 by Leave a Comment