The utility and energy industries haven’t really caught on to the whole cloud-hosted, software-as-a-service thing -- at least not to the extent seen in telecommunications, or industrial automation, to name two roughly analogous industries. And when utilities do pick up a new set of software -- something novel like cloud-based analytics or real-time IOT, not just an upgrade of some aging CIS -- they do so very slowly, with a series of steps between early pilots and revenue-generating deployments that can stretch from several years to nearly a decade.
Over the eight years or so that we’ve been covering the space, we’ve seen billions of dollars put into software startups promising to reinvent the way utilities do business, whether by hosting traditional utility software in the cloud, building new advanced data analytics applications that expand the utility’s reach, or deploying intelligent nodes into the grid itself. C3 IoT, the company formerly known as C3 Energy, has raised $1.4 billion on its own, and giants like GE and Siemens have no doubt invested heavily in Predix and Mindsphere, their respective big-data analytics platforms.
This level of investment may have exceeded the market’s capacity to turn it into value, however. Last month, demand response provider and would-be energy intelligence software provider EnerNOC announced it was seeking alternatives, including a sale, of its suite of C&I customer software solutions, built through a series of acquisitions over the past five years or so. CEO Tim Healy conceded that growth hasn’t nearly met expectations, and revenues from its software line of business have shrunk over the past two years, with a further decline expected this year.
EnerNOC’s situation is unique, of course, but the failure of its software to catch on could be partly driven by the fact that the overall market isn’t that big. This report pegs the global utility software market at $3.1 billion, with Oracle, SAP and Microsoft leading the pack, followed by Itron, Opower, ABB, Finserv, Trimble and Aclara. Together these companies represent just under half of the market, although it’s worth noting that this list is conspicuous for the absence of companies like IBM and OSIsoft, and must be taken as a rough guide.
Utilities don’t turn over their software platforms quickly, and must contend with regulatory approval as well as internal economics to do so. And with the exception of the largest utilities, most lack the technical expertise on staff to manage complex software deployments themselves. That leaves the sector largely controlled by a set of longstanding utility partners.
Ben Kellison, director of grid research at GTM Research, notes that many traditional software deployments were based around the hardware purchases utilities were making, with GE providing its own SCADA system for its equipment, Siemens doing the same, and on down the list. Other types of software now in general use across the majority of utilities, such as GIS and outage management systems, have followed a similar pattern of adoption, with utilities taking a decade or so to move en masse to a few select, proven platforms from a handful of vendors, he said.
These dynamics tend to favor incumbent providers. Oracle has customer information systems and meter data management, Microsoft has utility business management software, and SAP has big data analytics. Itron and Aclara are metering companies that have broadened into various customer and grid-facing specialties, ABB is a grid giant with a huge utility software suite it got when it bought Ventyx for $1 billion in 2010, and Opower’s behavioral analytics and customer engagement suite of products are now part of Oracle, which bought the startup for $532 million last year. (This purchase price was less than half of Opower’s IPO price in 2014, by the way -- a snapshot of the rise and fall of its perceived value in the market.)
The incumbent advantage is also driven by the fact that investor-owned utilities are predisposed to invest in software that’s tied to capital expenditures they can rate base -- say, support for new grid equipment, or data centers for smart meter data. This odd preference for capital costs as opposed to operating costs can work against software offerings based in the cloud, although we have seen a few states start to consider alternative structures to deal with this misalignment.
Aligning the data on market opportunities and adoption curves
Market forecasts from GTM Research underscore the link between capital expenditures and software investments. A report from 2015 projected the U.S. utility customer analytics market would surpass $1 billion by 2018, enabled by the spread of smart meters, smart thermostats and other useful nodes of data collection. And a 2016 report predicts a cumulative $10 billion in investment over the next five years in analytics to turn floods of smart meter data into useful intelligence. Of course, much of this growth is set to go to incumbent players, or the startups that have already gone through the work to earn a place in the slow utility adoption cycle.
Last month, the Department of Energy released its Modern Distribution Grid Advanced Technology Maturity Assessment (PDF), a report with useful inside information on where U.S. utilities stand on the adoption of different grid hardware and software. It’s one of three reports coming this spring from DOE’s next generation distribution system platform, or DSPx, initiative, which was created by request from utility regulators in California, New York, Hawaii and Minnesota, the states that are most active in creating new regulations to meet new distributed energy challenges.
The report breaks down the utility adoption cycle into five phases. The first two are R&D and pilots, and operational demonstrations that “expose the technology to the rigors of production deployment.” The next phase, early commercial deployment, is defined as “fewer than five operational deployments and fewer than three vendors offering commercially available technology products.” The “mature deployment” stage has multiple vendors supplying their technologies to multiple customers, and the final stage, “obsolete and replace,” is self-explanatory.
The report’s adoption curves are based on the extensive input of utility and industry insiders working on the DSPx project, and apply to 17 different technologies, ranging from volt/VAR optimization to distributed energy forecasting. Many are equipment-centric, but a few apply to cutting-edge software efforts.
Here’s the adoption curve for operational analytics, for example, which shows that GIS and outage management systems (OMS) technology are in mature deployment, as are AMI data management systems. But broader analytics software -- a category that includes much of the market opportunity over the next half-decade or so -- has only seen limited early commercial deployments so far.
In the critical area of distributed energy resource (DER) forecasting, things are moving even more slowly, with load forecasting and customer DER adoption modeling still in the operational demonstration phase and DER forecasting in pilots.
As for software to manage all this distributed energy, demand response management systems (DRMS) are crossing from early to mature adoption, but the category of more advanced distributed energy resource management systems (DERMS) and microgrids is still in the operational testing phase.
All in all, utilities aren’t even that far along in integrating their advanced distribution management systems (ADMS) with their OMS, GIS and CIS platforms, something DOE’s report pegs as a prerequisite for managing distributed energy resources at scale. Fewer than one in 10 utilities surveyed had an ADMS plan in place today, and only 20 percent had plans to have one by the end of 2019.
Acquisitions vs. standalone: The challenge for utility software
The long timespan between early and mature commercial markets, and the pressures brought on by incumbents that are rapidly snapping up competing software products, makes for a tough market for traditional venture capital investors. Some software startups have found success in the utility field -- but most of them have found their exits in acquisition, not standalone success.
Opower is a relatively rare example of a company going public first. Many acquisitions are more private transactions, making it hard to determine how investors have fared in the process. Some have been quite lucrative for them -- last year, Bit Stew, an innovative device and system data management and analysis software startup that had raised $26 million in capital, was bought by GE for $153 million. Others, such as Siemens’ purchase of eMeter in 2011, provided lower multiples (according to our reporting at the time), but still met investors' expectations (according to the company's statements).
It’s worth noting that GE Ventures was an investor in Bit Stew, fitting a pattern of strategic investments followed by acquisitions by entrenched players in the utility field. GE has had a voracious appetite for software companies of late, while also acquiring the field intelligence and distributed energy management smarts of rival grid giant Alstom.
Meanwhile, some startups that have kept going on their own have now entrenched themselves in the full maturity stage of utility deployments. Data visualization and analytics software provider Space-Time Insight is a part of many full-scale utility and grid operator deployments. On the customer side of things, Tendril has been chugging along for more than a decade -- though not without its own near-death experiences. Other customer energy management and engagement startups have gone the acquisition route, as EnergyHub did when it sold to Alarm.com in 2013.
But very few startups have aimed to reinvent the utility IT infrastructure wholesale. C3 Energy, founded by software billionaire Tom Siebel, is trying something like this, and has landed big enterprise deployments with Exelon, as well as newer deals like January's enterprise analytics contract with Con Edison and Orange & Rockland Utilities.
But the decision to rebrand as C3 IoT last year highlights the trend away from utility-centric business models, and toward the more nebulous, but potentially far more valuable, field of connecting and managing devices that make up the internet of things. Big AMI vendors like Itron, Silver Spring Networks and Landis+Gyr are also expanding their software to encompass streetlights, traffic sensors, air quality monitors, and other devices, with an eye on markets that could augment, or even outpace, the slow utility sales cycle. This brings with it a whole new set of competitors -- a subject that deserves its own treatment, which we'll leave to another time.