The big news behind Opower's latest behavioral demand response program results — announced this morning — may be that little has changed in the past year.

The top-line numbers are these: Opower connected with 1.5 million customers, through 29 called demand response events, including 12 million direct communications, and saved utilities 3% to 5% off peak demand.

If the 3% to 5% figure looks familiar, it's because Opower got the same results in the summer of 2014. But for the company and the demand management sector at large, that consistency is a big deal.

“In the utility industry, it's important to remember there is a lot of skepticism for things that are new,” said Josh Lich, Opower's director of product marketing. “The most common question we get is, 'Are you reliable?' This is something we're very proud of – just to repeat the result. That's what we were going for.”

And that reliability – especially over multiple iterations – is a tough bar to clear for behavior-driven programs. Other studies focused on broad demand response programs with no price signal have shown poor results. A 2013 study examining California's Flex Alerts found no reductions in power demand related to the emergency calls for load shedding from the state's ISO. And a 2008 examination of the same program cited customer confusion over the timely nature of the state's demand reduction requests.

An academic paper last year titled “The Perverse Impact of Calling for Energy Conservation” actually found no peak load reductions – and shoulder increases – in response to mass media calls for conservation. Researchers from the University of Tennessee wrote:“The evidence that consumers increase energy consumption after being exposed to conservation calls, yet no reduction in generation is observed over critical superpeak hours, is consistent with an emergent body of work in environmental economics exploring the unintended consequences of various policy actions.”

So for Opower, the stakes were high going into this summer. In addition to expanding its customer base by about 50%, the company also deployed its product internationally for the first time, working with Hydro Ottawa in Ontario, Canada. 

“We knew at this point, that it would be difficult to continue to drive additional impacts above and beyond the 3%,” said Opower's Nick Payton, associate director of product marketing and strategy. “We thought a challenge we could meet was to drive continued 3% savings in the face of customers who come in, perhaps disenchanted with receiving communications.”

“We wanted to test if we could create habits with customers that led them to continue to save.”

Closing the loop

While some research questions the efficacy of demand response calls absent a price signal, Opower has managed to consistently show a 3% reduction (the 5% peak reduction goes along with a peak savings rebate program). The key, says Payton, is personalization.

“We harp on it all the time,” he said, almost sheepishly. And it's true – the word turns up twice alone in the company's announcement this morning. But Opower executives swear it is the key to getting consumers to respond to demand response calls.

That lack of personalization – not the price signal – is the common failure in some demand response programs, said Payton. Blanket radio ads, press releases and social media calls all target broad groups with little context or comparison. Opower's communications show customers their energy use and response relative to neighbors and similar homes, essentially nudging them into keeping up with the Jonses.

In addition to calling for reductions in customers' power use, Opower sends a second communication within 24 hours of the event, effectively closing the loop.

“We're providing the feedback, we're giving them a sense of how the actions they took  had an impact,” Payton said. “It actually leads to habit formation, and for us that's a really big finding. It suggests this is a reliable product utilities can use in market. … The implication of persistence means this can be used more as a resource for planning.”

Room to grow

Behavior- and analytic-driven demand response programs, as well energy efficiency programs, are a growing market. Their scalability and software-based nature mean they can be deployed quickly, without the need for special equipment installations. But that flexibility comes with a price, said Navigant Consulting senior research analyst Brett Feldman.

“It's bit of a barrier on the utility side, the level of reliability of these programs, when you don't have direct control,” he said. “The trust is going to have to be gained as these programs proceed.”

But it appears that is happening. “We have looked at a lot of these types of programs over the past several years, a small base of results but it's growing every ear as the programs expand beyond the pilot phase,” Feldman said. 

Energy efficiency and demand response programs based around behavior and analytics can save from about 1% to 3%. “We are starting to build some confidence,” though more measurement and evaluation is in place on the energy efficiency side of things, Feldman said.

Navigant expects the sector to grow over the next decade, and that growth could accelerate if behavioral load management programs are allowed to bid into wholesale markets. Efficiency is already being bid in, and in some instances utilities are getting credit for behavioral demand response also.

“But you have to be able to prove you're getting savings,” Feldman said, which highlights the significance of Opower's latest results.

Wholesale is the Holy Grail

Different utilities are implementing Opower's BDR offerings in varying ways. Some are linking it with efficiency offerings, some using it as a traditional DR resource, and others are using it as a teaching tool – a first step towards getting customers to incorporate more technologies an savings programs.

“It can drive a broader awareness from customers,” said Opower's Payton. “They understand their electricity use in a broader context, and that actually helps prepare these customers to invest in new types of technologies.”

But the biggest interest is coming from customers looking to bid behavioral programs into wholesale markets, said Payton. The first of those is in the works and could be in markets by 2016.

“One of the strongest use cases we're seeing is using this as a resource that is relied on in a wholesale market,” said Payton. “We're looking forward to having this as a resource participating in at least one market next summer.”

BG&E and ComEd, with their programs tied to rebates returning closer to 5% savings, are already bidding into PJM capacity markets.

“In many ways they led the way in how to do this, and now there is a template,” said Lich. “You'd think you could take that template to the next utility, and they'd say 'Ok.' But it does take several years of proof points, often in their own territory, before they get comfortable.”