Big Data can be a lot like spring cleaning. You can come across a lot of "stuff" you don't really need, but you still have to dig into it. So where do you start?
If there's one thing that we know about Big Data-it's that Big Data looms "big" when it comes to determining just how you're going to cut into it to get something of value for your business. The task can seem formidable, and it's now causing serious heartburn for many C-level executives and middle managers.
Big Data can be a lot like spring cleaning. You can come across a lot of "stuff" you don't really need, but you still have to dig into it.
So where do you start?Finding business intelligence in Big Data depends on identifying strong key performance indicators (KPIs) that deliver high value to the business. For instance, if you're running a supply chain and are wondering which new vendors are likely to deliver on time and with quality, you won't find that data in your internal database (since the vendors are new), but you can find it on an Internet-based network shared with other companies that already have a history with that vendor. Here you can "mine" vendor on-time shipment data and you can go even further if you troll social media channels that can tell you something about this vendor's quality of work.
This combination of structured and unstructured data is immediately useful to purchasing and manufacturing-but it can be equally useful to marketing (which wants to reduce time to market risks from bad vendor performance), to customer service (which doesn't want followup warranty repairs) and to finance (which doesn't want any risk).
The points here are twofold:
- Big Data that answers critical business questions only "happens" when silos of data are broken down and mined—and the resulting data is combined into a composite answer to a key business performance indicator; and
- The investments a company makes in getting to this point only really work if the Big Data crosses enough silos of expertise within the company to uniquely integrate and summarize data.
To return to our example about on-time, quality vendors:
- Data silos were crossed when the data was mined from both external supply chain networks and social media.
- Initially, the company made the mistake of thinking that only purchasing and manufacturing would find this data useful-but when it thought a little more about the total key performance indicator built around vendor performance and its implications, the company also realized that the vendor information could tell marketing whether there could be risks to a major sales promotion if certain vendor were used; could tell customer service whether warranty followups based upon poor quality would be likely; and could tell finance what overall vendor risk levels were.
- As a result, when this Big Data was leveraged across the entire organization, organizational silos were broken down, and everyone shared in the benefit of the Big Data. Of course, the payoff of the Big Data was so much greater to the business.
All of this is important, because many enterprises today still remain fragmented in their use of information. Finance uses finance systems and marketing uses marketing systems. But Big Data investments only pay off "big" when everyone can exploit the data. This can require changes in organizational thinking on who "owns" and "can use" certain data.
- How can companies assure that their organizations "keep up" with the Big Data that they mine?
- Appoint someone as an overall "champion" or focal point for Big Data initiatives. This person should be tasked with getting different areas within the company to collaborate on Big Data KPIs and use.
- Never start a Big Data initiative without total top management support-from technology investments to recalibration of the organization for collaborative information sharing to collective definition of the key performance indicators that Big Data will be expected to provide answers to.