It’s not big data and utilities. It’s big data versus utilities.
This author no longer works for DNV GL.
“Big data and utilities” is one of today’s hottest topics in the utility industry. Oracle and eMeter have recently published two additional reports based on interviews with executives across the industry. Though their numbers vary, each indicates that less than half of utilities interviewed are confident in their current ability to handle big data analysis. Other industries, such as healthcare, retail, banking and telecom are already realizing huge benefits from leveraging BD—so why is the utility sector so far behind the curve?
It’s easy to point to the benefits: revenue protection, asset management/maintenance, outage management, and customer usage analysis regularly top the list. Likewise, the list of barriers to implementation usually read the same: lack of data analytic skills among employees, and/or insufficient IT hardware and software to perform the analysis are typically cited. Vendors promote cloud-based architecture and third party analytics packages as the solutions. While those can certainly help, I think folks are missing the fundamental problem. It is traditional utility organizational design and culture that are truly to blame. Look at how telecom (one of big data’s “poster children”) has transformed since the Ma Bell days—far more nimble, innovative, and responsive to customer needs and wants. Utilities, generally fighting deregulation at every step, still cling to the decades old, change averse model of delivering a commodity to “rate payers” for a fixed return.
A key component of BD analysis is enterprise-wide information sharing and benefits realization. Utilities’ heavily siloed (and often politically contentious) structures are not yet conducive to making that happen. Autonomous business units, typically not directly involved with advanced metering infrastructure (AMI) deployment projects, often have little understanding of the value that meter data can provide. The perception that meter data equates strictly to billing data is alive and well. Internally-led change management initiatives, hindered by the same lack of insight into the big picture of meter data’s business value, do little more than teach employees how to modify existing business practices to accommodate the new data format. Even the idea of retraining existing personnel to perform data analytics is problematic. Entrenched in the existing structure of separate sets of compartmentalized information, controlled by distinct business units, each serving specific needs, it’s hard to fathom these employees making the quantum leap to enterprise-wide information utilization.
The situation may look dire, but hope is around the corner. Within the halls of every utility sits a handful of enterprising analysts, typically buried in the load research, load forecasting, rates, or planning departments. These analysts, who usually operate under the radar of the IT department, have had to draw data from legacy billing and operations systems residing throughout the enterprise, then mate that data with data gathered from other external sources to build out their own analytics platform. They do this because they need to get their work accomplished. Think billing data coupled with square footage data publically available from tax records—now add NAICS or SIC codes gathered via a commercial survey or purchased on the open market, and we have ourselves a database that allows the analyst to do basic customer segmentation. Rudimentary yes, but over the years these individuals become the “go to people” within the utility enterprise when there’s a need for the information and data extraction skills they have available in their non-enterprise system.
Fundamentally, these individuals should be leading the big data charge at the utility. But the barriers, as mentioned above, can be daunting to the typical analyst who just wants to farm the data for information. In some cases utility executives have identified these individuals and have tasked them as primary business and functional contributors to the utility’s big data plan.