If you’re marketing hardware, software or services for “Big Data” – the analysis of very, very large datasets to uncover business opportunities – you should check out a recent column my colleague Larry Marion wrote for Datamation.
In it, he wrote that “despite a decade of expensive deployments and a parade of innovative products” customers are complaining that Big Data tools are too backward-looking and not predictive enough, can’t handle unstructured data, and are too slow and hard to use by non-technical, business types.
Solving these technical problems is up to the engineers, not us marketers. But the concerns in this chart provide a handy list of “pain points” for you to highlight in your marketing content, and in your social media searches for prospects.
Beer and Diapers? Naaah!
This report comes on the heels of a recent post I wrote for CA Technologies’ Innovation Today blog warning that organizations “won’t do the tough work of cleansing and validating (their data) to make sure the insights they gather will actually be valid.”
And last summer I covered a panel discussion that talked not only about data quality, but the very human factor that companies often don’t trust Big Data insights because they don’t fit their preconceptions. (Remember the oft-quoted Big Data insight that customers who buy beer often also buy diapers? Professor Tom Davenport of Babson College told the panel the convenience store chain never stocked the two together because it didn’t believe the sales data.)
While the database, analytics and hardware folks tackle things like speed, data quality, data access and usability, vendors can tap their internal subject matter experts to write about process issues such as:
- How does IT convince the business users of Big Data of the need to properly cleanse data, and then find the most cost-effective way of doing so?
- How can IT forge closer bonds with business units so it can help understand what are the Big Data questions most worth asking – and spending money to solve?
- Where are the hot new technologies in predictive analytics, which ones can be trusted, and what are cost-effective ways to try them out before putting big bets on their predictions?
Real or Hype?
All these are areas the sales force and Big Data consultants are probably already tackling, and have insights on, even if the engineering folks haven’t solved all the technical issues.
But all this is for naught if, despite all our marketing and posturing, Big Data is just not working for the vast majority of customers. Are concerns like those highlighted in this survey preventing, or just slowing, the insights promised by Big Data?