Uplight says opt-out enrollment took one utility's demand program to 70-80% participation

About eight in ten electricity customers say they want help managing their energy. About one in ten actually joins a utility demand management program. In a Utility Dive opinion piece published Nov. 11, Gisela Glandt, vice president of virtual power plants at Uplight, argued that the gap has little to do with interest and a lot to do with the sign-up process.

Her strongest evidence comes from one utility. A Midwestern utility that switched a program from opt-in to opt-out saw 70% to 80% of customers participate and grew its capacity four to five times faster, she wrote. A demand-side program manager trying to find megawatts before new generation can be built will want to know why that worked and where it stops working.

Why sign-ups stall near 10%

Glandt describes the opt-in process as a series of tasks for the customer. They have to understand the program, check whether they qualify, decide it's worth the trouble and then act. The reward can arrive months later, by which point they may have forgotten why they joined. Most customers drop out somewhere along the way, and she attributes that to effort, not lack of interest.

Uplight made the same case in a sponsored post on the Association of Energy Services Professionals blog in September. That post traces the 80% figure to an EY consumer survey and says enrollment has hovered near 10% for decades, despite bigger incentives, better devices and sharper marketing. Uplight's answer is what it calls built-in flexibility: participation that comes as a default setting or switches on when a device is sold, with no sign-up step. Where it has applied that approach, the company says participation rose from 10% into the 70% to 100% range.

Load growth is why this matters now. As Brian Martucci reported for Utility Dive, the Electric Power Research Institute expects data centers to account for up to 17% of U.S. electricity demand in 2030 and as much as 20% in 2035. Demand-side programs are among the few sources of capacity a utility can add in months. If they're stuck at one customer in ten, that advantage is mostly theoretical.

Two routes to default enrollment

Uplight separates two models. Behavioral demand response needs no hardware and costs the customer nothing, so it's open to almost every account, including renters and households without smart devices, which automated programs miss. Customers who respond to an event earn the incentive. Those who don't face no penalty, which one Uplight client summed up as all carrot and no stick.

The second model goes through device makers. A smart thermostat, connected water heater, EV or home battery is enrolled when it's set up out of the box, so it never needs a separate utility sign-up. Uplight cites ecobee's Grid Resiliency and Renew Home's Energy Shift Capacity as examples. The load shifts are smaller and steadier than a typical demand response event, and the AESP post says they're designed to go unnoticed by the household while covering more hours across a season.

That changes what a capacity planner is buying: less reduction per event, available over more hours. Uplight also points to a mismatch in the market. Smart thermostats are in 16% to 17% of homes, the highest penetration of any residential distributed energy resource, yet only one household in ten is enrolled in demand response. That suggests many thermostats already on walls aren't enrolled in any program.

Opt-out works only if quitting is easy

Glandt takes on the obvious objection directly: opt-out can sound like no choice. Her conditions are specific. Customers should be told early and plainly that they're enrolled, what that means and how to change it. Leaving should be as easy as joining would have been, through self-service tools and clear preference settings. And participation should be tied to outcomes customers care about, such as lower bills and a more reliable grid.

Under both models, customers can sit out a single event or day without consequence. That means the 70-80% figure measures how many enrolled households responded when dispatched, with an exit available at every step. A utility that defaults customers in but makes leaving hard is running a different program, and nothing in Uplight's figures speaks to how that one would perform.

What one household is worth to the grid

The per-home numbers explain why sign-up rates matter so much. Ontario's Save on Energy Peak Perks program, run by the Independent Electricity System Operator with EnergyHub, has enrolled more than 300,000 homes with smart thermostats and delivers 200 MW of peak load reduction, Charles W. Thurston reported in Solar Builder. Dividing one figure by the other gives about two-thirds of a kilowatt per home.

A program that needs 300,000 thermostats to deliver 200 MW depends on how many homes it signs, and the enrollment form decides that number. At that rate, moving from 10% to 70% participation in the same eligible population would multiply the megawatts roughly sevenfold, on our arithmetic, before anyone touches the incentive budget.

Batteries are a different case. In Commonwealth Edison's territory in PJM, an average residential solar-plus-storage system cuts peak demand by about 6 kW, consultant Scott Vogt told Herman K. Trabish of Utility Dive. That isn't a like-for-like comparison with Ontario's thermostats, but it shows how widely per-home value varies with the device. Large buyers have started paying for distributed capacity directly. In June, Google agreed to fund a three-year, 100-MW virtual power plant in PJM operated by Voltus, with Voltus responsible for meeting Google's capacity obligation.

That 100 MW is small next to PJM's record peak of 168.2 GW, set on July 2. A 2025 Department of Energy study found distributed resources could meet as much as 20% of U.S. peak needs by 2030. At under a kilowatt per thermostat home, reaching that scale will require enrolling most eligible homes, which is the problem Uplight is trying to solve.

The case for batteries first

Not everyone thinks thermostats are the right foundation. Geoff Ferrell, sonnen's senior vice president for its commercial and VPP project business in the Americas, told Solar Builder that thermostat and HVAC programs are essentially demand response with limited long-term returns. In his view, a permanently connected home battery is more dependable, because thermostats, EVs and other devices serve other purposes and may not be available when the grid needs them. It's a serious objection. Default enrollment can grow the number of homes in a program quickly, but each home still contributes very little.

Uplight's response is that low-effort programs feed the higher-value ones. At one Northwestern utility, 19% of behavioral DR customers went on to join thermostat, battery or EV programs. Among customers who leave an OEM-based program, 40% do so to join automated demand response. Read that way, the two views can both hold: default enrollment builds the customer base that battery and managed-charging programs recruit from. One caveat applies to the whole argument. Every result here is Uplight's own figure from unnamed utilities, published in the company's own commentary.

Running several programs at once

Layering programs creates its own work. Uplight's AESP post recommends cross-referencing enrollment when scheduling events, so a household in both a behavioral program and a device program doesn't get conflicting signals and its incentive is credited to the correct program. It also calls for a single management system that can view, forecast, dispatch and evaluate behavioral and device-based resources together.

This matters most for utilities that already run a bring-your-own-thermostat program and are considering behavioral DR in the same territory, where the same customers would show up in both. A program manager weighing a default-enrollment design could start with these questions:

• How many eligible customers would a default design reach, compared with today's enrolled count, and what does that difference add up to in megawatts at a realistic per-home value?
• Can the dispatch platform see a household enrolled in two programs and credit its incentive to only one?
• Does the program's measurement method count smaller, longer load shifts across a season, or only the reduction during a single event?