Smart thermostats do not save energy simply because they are smart. They save because they force homeowners to do the boring things: turn the temperature down, set schedules and stop heating an empty house. The official numbers behind that claim are modest and specific, and they deserve to be separated from the marketing gloss that surrounds connected heating hardware.
- Every degree counts: Raising the setpoint by one degree increases consumption by roughly 7%, according to Spain’s IDAE energy agency.
- Modulation matters: A thermostat that modulates boiler output can add around two efficiency points over a simple on/off unit, but only if both devices speak the same protocol, typically OpenTherm.
- Central heating caveat: In buildings where heating costs are split by floor area rather than metered consumption, lowering your own thermostat will not reduce your bill at all.
What the official figures actually say
There is no official savings figure for “smart thermostat” as a product category. What exists is public data on the behaviours a thermostat makes easier to sustain. The IDAE’s Practical Energy Guide, published in 2010 but still physically valid, offers three useful reference points: each degree of extra temperature adds about 7% to consumption; thermostatic radiator valves or programmable thermostats deliver savings of 8% to 13%; and 19 to 21 °C is enough for most people, with 15 to 17 °C sufficient in bedrooms at night.
The same guide recommends dropping the thermostat to 15 °C when you leave the house for a few hours and by 4 or 5 °C during longer absences. That is precisely what a connected thermostat automates: noticing you are out and lowering the temperature without you having to remember.
The honest reading is that if a smart device persuades you to move from 23 °C to 21 °C, the savings come from those two degrees, not from the app. Anyone already disciplined with a classic programmable thermostat will see only a small gain. Round savings claims printed on retail boxes should be treated with suspicion, since they depend heavily on how a home was heated before and cannot be extrapolated directly.
OpenTherm and the European efficiency classes
Modulating control does add a measurable slice of efficiency, and the European Commission has quantified it. In its 2014 communication on energy labelling for heating appliances, temperature controls are divided into eight classes, each with a contribution to seasonal efficiency. A simple on/off thermostat sits in class I with a 1% contribution, a TPI cycling thermostat with an on/off boiler reaches class IV at 2%, and a modulating thermostat paired with a modulating boiler reaches class V at 3%. Weather compensation with an indoor sensor climbs to class VI at 4%, while a multisensor control with three or more sensors and a modulating boiler tops the table at class VIII with 5%.
The gap between an on/off thermostat and a modulating one is therefore around two points. To capture it, the thermostat and boiler must communicate in the same language, which is where OpenTherm comes in. The OpenTherm Association describes it as a manufacturer-independent communication system between modulating climate appliances and room thermostats, designed primarily to control modulating boilers. If both devices support it, the thermostat can request “a little heat” instead of firing at full power. Check the boiler manual or ask the maintenance technician before buying.
Which heating systems can actually use one
Compatibility depends on how the heating is controlled today, and not every system accepts the same approach.
- Individual boiler with water radiators: the simplest case. A smart thermostat replaces the existing room thermostat, wired for on/off control or, where supported, OpenTherm.
- Heat pump (aerothermal): the IDAE highlights its efficiency, transferring between 2 and 4 kWh of heat per kWh of electricity. Control varies widely by manufacturer, so confirm with the installer which thermostat type the unit accepts.
- Electric radiators and convectors: these are independent units, each with its own element, so a wall thermostat cannot govern them alone. Remote control requires connected radiators or a system compatible with the specific model.
The central heating problem
In a flat with central heating, a smart thermostat only pays off if you are billed for what you consume. The relevant rule is Royal Decree 736/2020 on individual consumption metering in buildings with central heating. It requires individual meters where technically feasible and cost-effective, or heat cost allocators on radiators where those are cost-effective instead. In such cases each dwelling must have the means to control its own consumption, such as thermostatic valves on every radiator or zone valves with a thermostat. The variable portion of the bill, the part driven by usage, should sit between 60% and 75% of the total. Some buildings are excluded on technical or climatic-zone grounds.
The IDAE estimates that individual metering enables savings of 20% to 30%, but that saving comes from paying for consumption, not from the device itself. The practical consequence is blunt: if your community splits heating costs by coefficient or floor area, lowering your own temperature will not cut your bill, and a smart thermostat will not save you money. Where allocators or meters exist, the lever you can actually pull is the valves on your radiators, and connected thermostatic valves exist for exactly that purpose.
Smart thermostats work best for households that managed heating poorly before. For someone who left the system running at 23 °C all day, the effect may show on the bill; for someone who already programmed carefully, it is mostly comfort. The sensible starting point costs nothing: drop a degree, set schedules and bleed the radiators. If you still want phone-based control afterwards, treat the thermostat as a comfort upgrade with a modest efficiency bonus attached.