Sunday, January 20, 2013

Big Data is Not a Thing

Is Big Data a thing or an attitude?

We often read stories that use Big Data as a proper noun, either as representative of the future - "Big data: The next frontier.." or just something to deal with - "Is Big Data too big to back up?"   But as a noun, we run the risk of trying to characterize it with adjectives we understand and verbs we commonly use.

The opportunities provided for by the recognition and manipulation of streaming events (even if they're static for the moment) are quite new.   We now have the opportunity to work at the most granular events of processes rather than with surrogate aggregates.   Early in my career I worked with data from 300 customers and monthly warehouse withdrawals as a means of predicting new product success.  Today, the volume, velocity and variety of the available data allows for all kinds of nuanced simulated test marketing models.  All of which provide a better understanding of how and where consumers decide to try a new product.

Maybe we should look at the burgeoning field of data journalism - the purpose of which is to ultimately tell a story.  Mirko Lorenz depicts the process as follows, where data by itself has no real value.

Looks like a strategic planning or creative brief to me; just change 'public' to 'consumer'.   In most organizations, marketers are in the best position to do the filtering and visualization aspects - and those are activities not things.  

The early adoption of this approach is likely to be to confirm the stories we're currently telling - Do the data support what we claim?   The big data phase will be to create new stories from the ground up.  And it is there that game-changing stories will be found.

No comments: