The Splunk guy gave us the rundown on how CIOs should mine machine data for business intelligence. Hadoop has definitely grown from its origins in clickstream capture. The original definition of Big Data focused on its volume, velocity, and variety. Splunk adds "variability" to that mix while other data practitioners prefer to add "veracity." Landlords are using data streams from HVAC systems and building traffic to predict occupancy.
The panel's insights into Big Data prove to me that privacy is gone forever. The Facebook guy described how corporate HR profiles now capture manager effectiveness, attrition predictions, and school sources of the best hires. Everything that can be collected, will be collected. It all becomes part of our permanent records to follow us wherever we go in life. A lot of marginal performers are going to be permanently underemployed once even the slowest government bureaucracies harvest Big Data from their HR systems. The businesses that can't afford good ERP solutions or don't use their SaaS solutions like Salesforce to the fullest extent will be permanently stuck with low-quality data, inefficient supply chains, and poor employees.
Panelists did recognize the importance of Big Data in crossing functional silos and discovering duplication of effort. My readers know that I now take decision management very seriously. I say CKOs need to own the DM center of excellence in an enterprise to ensure that teams can draw data from anywhere. That is one way to fix the oft-heard criticism that CRM data is hard to link to the rest of the ERP system. I did not consider external traffic data to be significant until the panel remarked that it includes user downloads of software and documentation. I had previously considered the Small Data from product managers to be the most immediately useful stream because it included sales, returns, and unit costs. Their comments about external traffic made me think back to the time a few months ago when I downloaded a white paper from IBM and I had to log on to their site to complete the download. My contact info is still floating around in IBM's piles of casual contact records.
Here's my final observation, linking this Big Data stuff to my other recent discoveries. The cross-functional nature of the kind of Big Data opitimization this panel discussed is the reason a CIO cannot also be CKO, at least in very large organizations. The KM effort requires a cross-functional effort that touches domain users and the CIO's analytics teams need a KM intermediary when their work crosses paths with the domain users in product groups. The good news from the panel is that Salesforce modules now have built-in business intelligence dashboards. Domain experts in business groups no longer need to call IT every time they need to build BI tools.
I plan to attend more of these social functions from the various chambers of commerce and business groups representing international interests. There were in fact some model-quality Scandinavian women in attendance at this vikings event. Everything you're heard about the beauty of tall, blonde, Nordic women is true. That's reason enough for me to return. I can't wait until the Swedish Bikini Team gives a talk at Silicon Vikings.