This week I was able to spend some time at an American Marketing Association roundtable, on the future of data analytics. Lots of good people there and it was a valuable insight into data analytics applied to different business function.
Many of the models and analytical techniques are very similar, the difference simply being a shift in the problem being solved for. Ie will this consumer purchase (y/n?) vs will this consumer default (y/n?) and all the variations in between. It was very familiar.
Plenty of discussion around unstructured data and social media. Social media, unstructured data analysis and the ‘Big Data’ it creates is clearly the current ‘buzz’.
Although this topic is intellectually interesting it was also clear people have day jobs. I felt this tempered excitement a bit. After all in the real day to day world, targets need to be met, acquisition campaigns built and new business won.
The discussion actually felt a bit like the early days of the internet… everyone knows it will be a big opportunity, but it is not quite there yet and there are more pressing immediate priorities.
Big data is of course not information, but everyone has the sense there is value there… if only we could easily create structure, valuable information and insight could be extracted….
I do believe this is a huge opportunity, however just like the early days of the internet, we may need to think about this in a new way.
The current focus is on collecting as much information as possible (often unstructured) and then use techniques to create meaningful structured data. There is a lot of noise to eliminate and complexity to overcome in order to process…. it is the proverbial ‘needle in a haystack’.
However as the science develops a more efficient approach will evolve. Rather than analyze data after the fact, techniques are needed to create more useful (structured) data upstream and at point of creation.
Just as in the internet example, it will mean embedding data gathering intrinsically as part of day to day business, with process redesign to help facilitate. Data will be an output, not a by product, of any process.
The issue is this is difficult implement and any data gathering needs to be seamless for the consumer. It will require expertise and investment….. but for the pioneering companies that get this right, the volume of quality of data will be dramatically improved.
Targeted marketing to a degree we have not previously seen will be created….. a new breed of companies will be born.