Tiny Data + Unstructured Data = Big Data

mickyates Business, Change, Customer, ideas, Innovation, networks, Social Networks, Strategy, Technology Leave a Comment

I gave a presentation at the President’s Lecture of the Chartered Management Institute a couple of days ago, on behalf of dunnhumby, on the business and organizational change implications of “Big Data“.

There seems to be two fundamental strategies to use the insight from “Big Data” and turn it into useful insight and business decisions – Customer Centricity and Innovation Networks, both of which I have written on before – and will do so in future.

One point I would like to stress here, though, is the definition of “Big Data” itself. There wasn’t one that I could find that proved completely satisfactory, as most address the technology aspects and challenges and ignore the organizational implications. So I offer this:

“Big Data” is
  • complex: comes from multiple sources – structured databases and unstructured social
  • analysable: it must be captured, processed, analysed & visualised
  • useful: insight must create decisive action
  • pervasive: it impacts everyone – changes everything in the organization’s processes

To look at this another way, I offered this summary.

Tiny Data + Unstructured Data = Big Data

Tiny Data means data from a single source in a structured format which, whilst it may in a huge quantity, is actually limited in its complexity. The CEO of Visa Europe, Peter Ayliffe, is the President of the CMI – and he agreed that even the databases held by Visa are, with this definition, “tiny”.

Unstructured Data means exactly that – no fixed database format or coherent structure. Think of messages sent on Twitter, images uploaded to the web, Facebook posts and likes, phone calls, customer service calls and so on.

“Big Data” combines the two. Only now are technologies becoming available to combine and make sense of these different sources – and most importantly turn the analytical results into useful insight and action plans.

I gave the example of looking at someone’s Facebook timeline, and noting that they tend to like wearing blue but never orange. If you are a clothing manufacturer, and knew that fact, wouldn’t that help you make more appropriate offers to that potential customer? And if you could match this insight against the customer’s purchase records over time, wouldn’t that give a richer insight into their behaviour?

Of course, it raises privacy issues. I will explore this and related ideas in future posts, and in particular the Leadership and Organizational Change implications of this “Big Data’ revolution.