Cleantech Market Intelligence
Turning Point for MDM
“When you come to a fork in the road, take it.”
That sublime Yogi-ism mirrors where meter data management (MDM) vendors find themselves. Will MDM become the data analytics of choice for utilities, or will it become a piece of critical yet mundane middleware, relegated to managing meters? And will MDM vendors have any say in their destiny? Understandably, they would like to believe that they do. So they are not hanging about, waiting to see what happens.
One year ago, Pike Research published its report, Meter Data Management. In researching the forthcoming 2012 version of this report, I’ve found quite a bit of movement since last year – much of it aimed toward the vendors’ continued existence. The two most obvious trends are:
- Acquisition: Nearly all of the major MDM vendors have been acquired. Headlines include Siemens acquiring eMeter and Landis+Gyr acquiring Ecologic Analytics.
- Data Analytics: Everywhere one looks, MDM vendors have repositioned themselves as data analytics vendors. Some have stopped talking about MDM and instead offer MDA – Meter Data Analytics. A year ago this was a green shoot; now it is in full bloom.
This near-unanimous repositioning of MDM vendors as data analytics vendors is understandable. My favorite hype indicator is webinar announcements. Today nearly every webinar invite talks about some version of “unlocking value through data analytics.” If you’re an MDM vendor and your alternatives are to stay the course or to surf the analytics wave, the decision is obvious.
Unfortunately, the path from transaction-based MDM to analytics engine is much less obvious. In extreme situations it may not be possible at all. My colleague Carol Stimmel has recently blogged that utilities may make use of a great deal of data – some quite lacking in structure – to understand and influence markets. (For more detail see her forthcoming report, Smart Grid Data Analytics). It’s tough to imagine how an MDM system could digest unstructured data such as demographics and weather.
On the other hand, there is a middle ground where clever analysis creatively correlates transactional data only. Survey results often do this. Either MDM or an analytics engine could make those correlations. But which will prevail? I suspect that MDM will get those tasks, for two obvious reasons. First, utilities are notoriously conservative and are likely to prefer the existing known quantity called MDM to something new and unknown. Second, using MDM requires no new, expensive, and possibly disruptive software deployments. However, MDM vendors may overreach into the world of unstructured data analysis. Such a move could consume enormous amounts of resources for a low probability of success.
No matter what happens, there are certain features of MDM that will remain indispensable. Validation, estimation, and editing (VEE) is essential for creating a high quality system of record for metering data. VEE must process complex business rules over a massive data set, quickly. This is where the transactional orientation of MDM remains an asset. Likewise, utilities depend upon MDM to create ever-more complex billing determinants. Such computations are not appropriate for an analytical engine. No matter how much turf analytics ultimately capture, MDM will remain essential at every utility.
Perhaps the most logical summary of where MDM will end up vis-à-vis analytics is “horses for courses.” MDM systems do lots of things that analytics engines cannot do, and vice versa. And regardless, MDM vendors have survived years of being thrown into AMI deals as a no-cost extra. Fighting off upstart analytical engines should be child’s play!