Last September, in the lead up to its public offering, Opower (OPWR) announced its effort to reinvent residential demand response through the use of behavioral science.

"This is game-changing," said Roderick Morris, Opower's senior vice president of marketing, after the company cut residential electricity demand for Baltimore Gas & Electric (BGE) during peak times by 5 percent simply by sending customized emails, text messages, phone calls and mailers to homeowners.

Naturally, the company was enthusiastic about the pilot results. But when Opower filed its S-1 months later, it included some cautionary language about the uncertain future for behavioral demand response.

"We have invested considerable resources in developing a demand response offering. While the early results have been promising, it is possible that the market could be smaller than we have expected or that our product will not function as intended," wrote the company.

Outlining the risks associated with a company's strategy is standard in an S-1. But the admission proved the difficulties companies like Opower face when trying to sell new energy management products to conservative utilities.

For now, however, it appears that Opower's behavioral demand response program is functioning and expanding as intended. This morning, it announced a deal with BGE to scale the program to 800,000 customers in Baltimore, up from last summer's 300,000-participant pilot. Opower has also signed up three new customers: Consumers Energy, Efficiency Vermont and Glendale Water & Power.

The success of BGE's behavioral demand response pilot has helped Opower attract new utility customers. During four peaking events last summer in Baltimore, three-quarters of homeowners saved an average of $8 per event, said Chris Burton, BGE's vice president of smart grid & technology.

For the 100,000 customers with direct load-control switches on their air conditioning units, BGE and Opower saw a 23 percent reduction in energy use. For the other 200,000 customers with no devices installed, they saw a 5 percent reduction. "The behavioral demand response capability made the system more reliable at the very peak time," said Burton.

Based on the results with BGE, Opower claimed it could expand customer interaction twenty-fold and lower the per-kilowatt cost of residential demand response by 40 percent compared to traditional direct load control on air conditioning units.

Kevin Hamilton, a senior director at Opower, said that utilities are more willing to consider the offering now that the concept has been proven.

"Even for utilities that have been doing direct load control, this now gives them access to the other 75 percent of customers they weren't interacting with," said Hamilton. In BGE territory, 85 percent of customers were aware of the demand reduction events, versus 40 percent customer awareness levels during similar load control programs in California.

BGE had two factors in its favor that proved critical to the success of the program: smart meters and time-of-use pricing. The smart meters helped provide more real-time information on customer usage, and the pricing framework encouraged more people to shift their demand. Many utilities don't have either.

Hamilton said Opower is attempting to prove that it can scale behavioral demand response without those two factors.

Opower has been talking with engineers and economists about how to measure the impact of its behavioral messaging without smart meters. One option is to measure reductions on feeders serving thousands of customers at the substation level. 

"We have reason to believe that we can produce savings without rates or meters," said Hamilton. If Opower proves that its version of demand response works in a variety of markets with different structures, it could move well past the milestone of 1 million customers by next summer.

The company is banking on this new offering as a way to expand its reach in the utility sector. In its first-quarter earnings report last week, Opower reported a net loss of $7 million. The company doesn't plan to be profitable until 2017 as it expands new products and builds momentum internationally.

"You will see widening loss. That is the plan. We’re doing that to beef up our sales and marketing investments, as well as being able to invest in continued research and development in our product set," explained CFO Thomas Kramer.

To hear more from Opower, join Greentech Media at Grid Edge Live in San Diego June 24-25, where CEO Dan Yates will join our fireside chat.