You Spent $15,000 on a Heat Pump. Your Thermostat Is Running It Like a $40 Space Heater.

Small orange letters spell "auxiliary" on your thermostat screen and stay for eleven minutes, then twenty, then they don't go away at all, and your January electricity bill arrives at $387 instead of the $160 your installer promised when he sold you a system with a coefficient of performance of 3.8.

You paid for a machine that moves heat, and your thermostat decided, on its own, without asking, to make heat instead, using the most expensive method available to it, because no one at Google or Ecobee or Honeywell changed the factory default that assumed you still had a gas furnace.

The machine and the middleman

A heat pump with a COP of 3.0 delivers three units of thermal energy for every one unit of electricity it consumes. It does this by moving heat from outside air into your home, even when the air outside is cold, because thermodynamics is more generous than intuition suggests. A 10 kW heat pump drawing 3.3 kW of electrical power produces 10 kW of heat output. That ratio is the entire economic argument for replacing your gas furnace.

Electric resistance heat strips bolted inside the air handler next to that heat pump have a COP of exactly 1.0, which means one kilowatt in and one kilowatt out, no thermodynamic leverage, no heat transfer from an external source, just a coil of nichrome wire getting hot with the same physics as the $40 ceramic heater you bought at Target for the garage.

When those strips activate, your 3.0 COP system drops to 1.0 on the portion of heating they handle. If the strips are providing half the heat output, your effective COP is about 1.7. If the strips take over entirely, you are paying three times more per BTU than your heat pump would have cost you, and the $15,000 compressor sitting outside your house is doing nothing.

Whatever device decides when those strips fire is your thermostat, and your thermostat was not built for this job.

Designed for a different machine

Nest, Ecobee, and Honeywell all designed their flagship smart thermostats around a simple mental model: the homeowner sets a temperature, the system either heats or cools until it reaches that temperature, then it stops. This model works beautifully for gas furnaces, which produce 120°F supply air, ramp quickly, and have no auxiliary backup system to accidentally engage.

Heat pumps are different machines entirely. A variable-speed heat pump modulates its compressor continuously, producing lower-temperature supply air over longer run cycles, and it prefers to cruise rather than sprint. When the thermostat detects that the indoor temperature has drifted more than two degrees below the setpoint, its firmware concludes the heat pump cannot keep up and calls for auxiliary heat strips to close the gap. This is the default behavior on most smart thermostats, designed to prevent discomfort, but what it actually prevents is efficiency.

A Lawrence Berkeley National Laboratory review of smart thermostat and heat pump compatibility found that most third-party smart thermostats use step-level staged control, while variable-speed heat pumps require a deeper integration that respects the lower temperature and longer cycle times of heat pump operation. Proprietary thermostats that ship with heat pump systems offer that integration, but they are typically harder to use and lack the learning algorithms, remote access, and energy reporting features that drive consumer adoption of Nest and Ecobee, which means homeowners face a choice between usability and compatibility, and most of them do not know the choice exists.

Net result: homeowners install a $250 smart thermostat to save energy, and the thermostat's factory defaults increase their energy consumption by engaging expensive resistance heat they don't need.

The default settings and what they cost

Three settings determine how much unnecessary auxiliary heat your system uses, and most homeowners have never changed any of them because the settings are buried in installer menus that require codes or multi-step navigation to reach.

Auxiliary heat threshold. This is the temperature differential between the setpoint and the current indoor temperature at which the thermostat activates the heat strips, and many thermostats default to a dangerously aggressive 2°F. That means if you set your thermostat to 70°F and the house is at 67.5°F when you wake up, the strips fire immediately instead of letting the heat pump recover on its own. Setting this to 4°F or 5°F eliminates most unnecessary auxiliary heat engagement in mild to moderate climates.

Compressor lockout temperature. Below this outdoor temperature, the thermostat shuts off the heat pump compressor entirely and runs only the strips, and some thermostats default to 35°F or 40°F despite the fact that modern cold-climate heat pumps operate efficiently down to 5°F and continue producing useful heat at -13°F. Trane demonstrated a unit running at full capacity at -15°C (5°F) and continuing to produce heat at -30°C (-22°F) during the Department of Energy's cold climate heat pump challenge. If your compressor lockout is set at 40°F, you are paying resistance-heat prices for every degree below that threshold, even though your heat pump has another 35 degrees of headroom.

Setback recovery behavior. A gas furnace can recover from a 5°F overnight setback in fifteen minutes, but a heat pump working at lower supply temperatures might take forty-five, and if the thermostat interprets that recovery time as a failure to maintain temperature, it fires the strips for the entire recovery period. One HVAC contractor writing on GreenBuildingAdvisor documented cases where thermostats were wired incorrectly, activating auxiliary heat strips during cooling mode, so the system was simultaneously heating and cooling the same air while the homeowner wondered why their electricity bill looked like a data center's.

Adjusting these three settings properly can reduce winter heating bills by 20 to 40 percent on homes with heat pumps, no new equipment, no service call, just navigating three menus that most homeowners don't know exist.

What the lab already solved

Researchers at Purdue University's Ray W. Herrick Laboratories ran a field demonstration of predictive heating control on an occupied, 2,240-square-foot single-family home in Indiana during January through March 2023, with outdoor temperatures dropping as low as -15°C (5°F) and the home running an air-to-air heat pump with electric resistance backup, exactly the configuration in millions of American homes.

Instead of reacting to temperature gaps after they appeared, the predictive control system adjusted indoor temperature setpoints based on weather forecasts, occupancy patterns, and data-driven models of the building envelope and heating equipment, anticipating demand before the house got cold enough to trigger emergency measures.

Over the test period, daily heating energy consumption fell 19 percent (95 percent confidence interval: 13 to 24 percent), energy consumed by backup resistance heat dropped 38 percent, and usage of the highest stage of backup heat, the 19 kW strips that drive the worst spikes in electricity demand, fell 83 percent. Concurrent surveys of the home's residents confirmed that thermal comfort was maintained throughout.

Annualized, the researchers estimated savings of approximately $300 per year, with a 95 percent confidence interval placing the reduction between 23 and 34 percent of total heating costs, and for a system that requires no hardware changes and runs on the same thermostat wiring already in the wall, that is a payback period measured in months, not years.

But the control system exists as a research prototype, and it is not available in any consumer thermostat sold today.

3.64 million reasons this matters now

Heat pumps outsold gas furnaces in the United States in 2025 by 11 percent: 3.64 million heat pump shipments versus 3.25 million gas warm-air furnaces, according to AHRI shipment data reported by ACHR News, a structural crossover that happened for the first time in history and is not reversing.

Federal incentives are accelerating that shift. Under the Inflation Reduction Act's High Efficiency Electric Home Rebate program, households earning below 80 percent of area median income qualify for up to $8,000 in point-of-sale rebates for heat pump installation, and all owner-occupied homes can claim a $2,000 annual tax credit under Section 25C with no income limit. Rocky Mountain Institute projects at least 12 million new heat pumps deployed by 2030, a floor estimate that does not fully account for state-level mandates and utility incentive programs in more than fifteen states.

Every one of those heat pumps will be paired with a thermostat. Most of those thermostats will ship with defaults designed for the gas furnace the heat pump replaced.

Why the fix hasn't shipped

Technical barriers barely exist. Purdue's system used weather forecast APIs, a learned building thermal model, and occupancy data that smart thermostats already collect, and Google Nest already has hardware capable of running this algorithm while Ecobee already has the occupancy sensors, so the computational requirements are trivial compared to the machine learning these devices already perform for scheduling and energy reporting.

Liability is what keeps the algorithm off the shelf. A thermostat that prevents auxiliary heat from firing in cold weather risks a scenario where the home's temperature drops below a safe threshold, pipes freeze, and the manufacturer faces a product liability claim, which is why conservative defaults are not accidents of engineering but decisions of legal departments: comfort is guaranteed while efficiency remains your problem.

Contractors compound this, because many HVAC installers do not configure thermostat settings after installation. A ServiceTitan survey of 1,000 residential contractors published in April 2026 found that 74 percent view AI as an efficiency tool, but only 25 percent are currently using any AI-driven optimization, so the thermostat goes on the wall with factory defaults, the installer leaves, and the homeowner spends the next five winters paying triple for heat they didn't need to buy.

What to check before next winter

If you have a heat pump with a smart thermostat, open the installer settings menu. On Nest, this is under Settings > Equipment > Heat Pump. On Ecobee, under Settings > Installation Settings > Thresholds. On Honeywell, under Installer Setup (ISU), which may require a contractor code.

Set the auxiliary heat threshold to at least 3°F, preferably 4°F or 5°F. If your heat pump was manufactured after 2020 and is rated for cold-climate operation, set the compressor lockout to the lowest available temperature or disable it entirely. Configure the recovery behavior to "gradual" or "economy" mode rather than "aggressive" or "fast," so the thermostat doesn't interpret a normal heat pump recovery cycle as a system failure that requires emergency backup.

Monitor your system for the first two cold days after making changes. If the house reaches setpoint within 90 minutes of a 5°F recovery without auxiliary heat, your settings are correct. If the house fails to reach setpoint after two hours, raise the auxiliary threshold by one degree and retest.

None of this is complicated work; it's a five-minute adjustment that should have been made the day the system was installed, and it will save you more money per year than the smart thermostat itself cost.

Limitations

Purdue's field study tested a single home in a single climate zone (IECC Zone 5A, cold-humid) over a single heating season, so the 19 percent energy reduction and 38 percent backup heat reduction may not generalize to homes with different insulation levels, duct configurations, or heat pump sizing. Homes with dual-fuel systems, where the backup is a gas furnace rather than resistance strips, face a different optimization problem with different cost tradeoffs that this article does not address. ServiceTitan's survey captures contractor sentiment, not measured field performance, and the IRA rebate figures assume current program structures that may change under future administrations. Savings estimates of 20 to 40 percent from thermostat setting optimization are drawn from manufacturer and contractor guidance, not from a controlled study with a baseline comparison group.

Against this article's thesis

Conservative thermostat defaults exist because real consequences follow when they fail, and a predictive algorithm that underestimates heating demand on a -10°F night risks frozen pipes, which cause an average of $12,000 in damage per incident according to State Farm claims data. Google and Ecobee are not ignoring the optimization opportunity; they are rationally choosing to let the homeowner bear a $300-per-year efficiency penalty rather than accept a low-probability, high-consequence liability risk that could cost their company far more, because for a manufacturer selling 10 million thermostats, even a 0.01 percent failure rate producing freeze damage would generate 1,000 claims per year. Seen through that lens, the defaults are not broken at all; they are insurance, priced correctly for the manufacturer if not for you.