Octopus Energy US gets 50% to 60% of smart-device owners into demand response

The historical benchmark for utility demand response programs is 5% to 10% participation. Octopus Energy US says it gets 50% to 60% of its customers with a smart device to take part.

The figure comes from Nick Chaset, CEO of Octopus Energy US and Uplight. He holds both titles because of the recent strategic investment that brought the two companies together. The idea is to combine the consumer-facing user experience Octopus built with the deep utility relationships Uplight holds across North America, and use the pair to unlock distributed energy resources at scale.

Behind the number sits a harder question for whoever owns a utility's flexibility program. Can household devices be called on often and reliably enough to count as a resource rather than a weather contingency? When demand was steady, Factor This notes, the inefficiencies of low-participation programs could be glossed over. Load growth from data centers and pressure on customer bills are pushing that question onto planning agendas, because new capacity is slow to build and expensive to pay for.

From fifteen calls a year to dozens

Demand response has traditionally been treated as a seasonal emergency valve. Traditional models have relied on top-down utility tools and broad email blasts rather than consumer-first engagement, and too often utilities rely on the stick without offering the carrot.

Those habits matter because willingness was never the scarce input. Factor This puts it at roughly 80% of consumers, which makes the single-digit participation history look like a design problem rather than an appetite problem.

Uplight General Manager Hannah Bascom describes a different operating rhythm. In her account, utilities can call on the portfolio of assets Uplight enrolls and manages in many kinds of circumstances, dozens of times a year, and she presents it as a multipurpose grid tool. Her goal, she said, is for Uplight to become the grid's Swiss army knife. When customer assets connect directly to a utility's core operating systems, the utility can shift away from passive energy efficiency programs and toward active, dispatchable load management, a resource it can rely on every day.

Octopus reports participation of 50% to 60% among customers with a smart device, compared with a historical benchmark of 5% to 10%. Chaset traces the gap back to the company's starting point. As he tells it, Octopus puts the customer's user experience ahead of everything else, and that is what lets utilities measure these value streams in a tangible way. For a planner, the word tangible carries weight, because a resource only counts if its contribution can be measured.

There's one limit on the headline figure. It counts customers who already own a smart device, and the reporting doesn't say whether those customers were enrolled opt-in or opt-out, or over what period participation was measured.

Factor This frames the scaling story as participation moving from a fraction of customers to daily routines involving more than half of connected households. At that scale, it argues, distributed energy resources stop being only a reliability tool and become a direct driver of long-term affordability. That is where the household pitch begins.

Selling flexibility on the monthly bill

Chaset's argument for why people enroll is simple. Nobody buys a smart thermostat or an EV to join a virtual power plant. They buy the EV because it's a better car, and the thermostat because the house is more comfortable. The utility's job, as he frames it, is to point out that the hardware a customer already owns can cut their bill.

He also calls demand flexibility the main short-term tool a utility has for bringing bills down while rates climb quickly, especially where a time-of-use rate is in place. In his telling, shifting about 10% of a household's consumption from peak to off-peak hours produces real savings. No dollar figure tied to a specific tariff came with that claim, so the size of the savings will depend on each utility's peak and off-peak price gap.

There's a regional precedent for leading with the reward. Opt-out behavioral demand response programs have been a win for customers in the Midwest, according to Factor This, which describes them as an "all-carrot" model that gives households agency over energy costs without touching their comfort.

That is the mirror image of the stick-first approach that kept participation in single digits. It also suggests that enrollment design deserves as much scrutiny as device counts when a utility compares one program's participation rate with another's.

Data centers as the paying customer

The bill argument gets sharper where data centers are arriving. As hyperscale load drives historic growth in regions such as Texas, Arizona and Virginia, local communities are pushing back against new infrastructure because they're worried about their bills. Bascom puts it bluntly: "People are increasingly becoming the blockers to data center development."

Her proposed answer runs through the same household devices. Giving those customers tangible benefits, such as lower bills and more comfortable homes, frees up the capacity data centers need to reach speed-to-power, she said, and it makes power more affordable for the utility and for its end customers alike. Seen this way, the flexibility fleet does two jobs at once. It delivers the bill relief that could soften local opposition, and it creates the headroom that could let new load connect sooner.

That opens up a funding model Factor This describes as one possible approach: data centers that need speed-to-power help pay for home batteries and smart hardware across a utility's service area. It's an idea, not a signed program. But if it took hold, a large-load customer would be paying to build a household flexibility fleet, and the neighbors who were worried about their bills would end up holding the savings.

For utilities whose communities are pushing back on new data center infrastructure, a demand response program could become part of how a data center gets connected, rather than a separate customer program running alongside. Factor This argues that making the link stick requires utilities to define and communicate how better-defined programs translate into local community benefits.

The benchmark is a 135.5 MW battery

Utilities already have another way to buy flexibility, and it comes with a spec sheet. Xcel Energy has taken delivery of battery units at its Blue Lake site in Minnesota, Factor This reported in a story republished by Power Engineering. The 135.5 MW facility will hold 132 battery units that can deliver power for up to four hours, and it's expected to start serving customers in 2027. Together with two planned projects in Becker, Xcel expects to run more than 735 MW of battery storage in Minnesota by the end of 2027, which it says will be the largest battery portfolio in the Upper Midwest.

The Blue Lake batteries will store energy generated across Xcel's Upper Midwest system, from wind, solar, nuclear and natural gas resources. Xcel argues that storage has evolved enough to be deployed at existing energy sites like Blue Lake, reusing infrastructure already in place. That is its own version of getting more out of what a utility already owns, the same appeal Chaset makes for devices sitting in customers' homes.

That build-out is part of a bigger wave. BloombergNEF projects 188 GW, or 535 GWh, of storage installed worldwide in 2026, with 62% of it in China, the United States and Australia. It expects the U.S. to deploy 30% more storage over the next decade than it forecast just six months earlier, reaching 213.9 GW cumulatively by 2030 and 434.4 GW by 2036. Global deployments could pass 3,098 GW, or 11,106 GWh, in 2036, roughly 10 times today's totals.

Bria Shea, president of Xcel Energy for Minnesota, North Dakota and South Dakota, describes storage as letting the utility hold energy and dispatch it when it's needed most. She pointed out that for over 100 years, power providers have produced electricity at the moment customers call for it and sent it out to them almost instantly. "Battery storage changes what's possible," she said.

Justin Meagher, Xcel's battery storage project manager, explains the job in terms that sound a lot like Bascom's. In his description, the systems react fast when conditions shift, bank energy while supply is there, and deliver power during periods of high customer demand. Both companies are selling dispatchable flexibility. The difference is what sits behind the dispatch signal.

Here's the strongest objection to the household-fleet argument. A battery's output is a physical rating the utility owns and controls. A fleet of thermostats and EVs delivers only as much as the households that stay enrolled and let the event run. On the other side of the ledger, customers paid for the devices themselves, for their own reasons. So the two options carry very different capital and certainty tradeoffs, and each needs its own way of being evaluated.

What would prove the model out

The claim worth testing is that participation holds once calls go from fifteen a year to dozens. If utilities working with Uplight can show enrollment and event response staying near Octopus's reported range under that schedule, household devices earn a place in resource plans alongside projects like Blue Lake. If participation slips as calls get more frequent, the program drifts back toward the seasonal valve it was built to replace.

The battery side of that comparison comes with a date. Xcel's Blue Lake battery system in Minnesota is expected to begin serving customers in 2027, according to Power Engineering. A household program that has a season or two of frequent-dispatch results in hand by then would let a planner compare the two options on measured performance.