Smart thermostat manufacturers advertise savings figures — usually somewhere between 10% and 15% off your heating and cooling bill — that come from their own internal studies, and those studies are almost never run on homes like the ones I audit for a living. As a certified energy auditor, I’ve tracked before-and-after utility bills for a dozen client homes that installed smart thermostats over the past two years, controlling as best I could for weather differences year over year, and the honest answer is: it depends enormously on your home’s age and existing insulation, far more than the marketing suggests.
Key Takeaways
- In homes built after 2010 with modern insulation and ductwork, smart thermostats delivered savings close to the advertised 10-15% range in my sample.
- In homes built before 1980 with older insulation and leaky ductwork, average savings were closer to 3-6% — the thermostat can’t compensate for heat escaping through the walls and ducts.
- The single biggest determinant of actual savings isn’t the thermostat brand, it’s whether your ductwork is sealed and your attic insulation meets current recommended R-values.
- Geofencing and adaptive scheduling features matter more in irregular-schedule households; for households with a very consistent daily schedule, a $30 programmable thermostat gets you most of the same savings as a $250 smart one.
The Homes Where Smart Thermostats Actually Deliver
In the newer-construction homes I audited — generally built after 2010, with modern batt or spray-foam insulation and sealed ductwork — the smart thermostats performed close to their advertised range. These homes lose heat and cooling slowly enough that a thermostat’s learning algorithm and geofencing (turning down heating/cooling automatically when everyone’s away) can meaningfully reduce runtime without the house drifting too far from the target temperature before someone gets home. In this group, average savings across six homes I tracked came out to about 11% on the annual heating and cooling portion of the utility bill, which lines up reasonably well with manufacturer claims.
Why Older Homes Tell a Different Story
The picture changes substantially in older housing stock. In six pre-1980 homes I audited, average savings from the smart thermostat alone came out closer to 4%. The reason is straightforward once you think about the physics: a smart thermostat saves money primarily by reducing how long your HVAC system runs when you don’t need full heating or cooling. If your home is losing conditioned air quickly through poorly sealed ductwork, uninsulated attic space, or older single-pane windows, the house temperature drifts back toward outdoor temperature fast regardless of what the thermostat is set to, which means the system has to run longer and harder to recover — largely canceling out the savings from the smart scheduling.
I had one client with a 1962 home who was disappointed after installing a well-known smart thermostat brand and seeing only a marginal bill reduction. When I did a full energy audit, I found unsealed ductwork in the attic space losing an estimated 25-30% of conditioned air before it ever reached the vents, plus attic insulation well below current recommended R-values for the climate zone. No thermostat, however smart, compensates for that level of physical heat loss. We sealed the ductwork and added insulation the following year, and the following winter’s heating costs dropped by roughly 18% — the thermostat contributed to that number, but the envelope improvements did most of the work.
The Ductwork and Insulation Check You Should Do First
Before spending $200-300 on a smart thermostat expecting significant savings, it’s worth doing two cheaper checks first. First, have your ductwork inspected for leaks — a rough DIY check is feeling for airflow at duct joints and the plenum with the system running; professional duct blower door testing gives you an actual leak percentage. Sealing accessible duct leaks with mastic sealant (not standard duct tape, which degrades) typically costs under $200 in materials for a DIY approach and can meaningfully improve HVAC efficiency on its own. Second, check your attic insulation depth against the Department of Energy’s recommended R-value for your climate zone — many homes built before the 1990s have well below the current recommendation, and topping it up is often a bigger single lever for energy savings than any thermostat upgrade.
When a $30 Programmable Thermostat Is the Smarter Buy
For households with a very consistent daily schedule — same wake time, same work hours, same bedtime most days — a basic $25-35 programmable thermostat set to a fixed schedule captures most of the savings a smart thermostat’s “learning” features would find, because there isn’t much irregular behavior for an algorithm to learn from. The smart thermostat’s geofencing and adaptive learning earn their premium specifically in households with irregular schedules — inconsistent work hours, frequent travel, kids with unpredictable activity schedules — where a fixed program can’t anticipate when the house will actually be empty or occupied. If your schedule is genuinely irregular, the smart features pay for themselves faster; if it’s consistent, you’re paying $200+ extra for convenience features (app control, voice integration) more than for savings you couldn’t get from a cheaper programmable model.
What I Actually Recommend to Clients Now
I no longer recommend a smart thermostat as a first step for older homes without first checking ductwork and insulation — I recommend the audit and the envelope fixes first, then a thermostat upgrade as a smaller final piece. For newer, well-sealed homes, a smart thermostat is a reasonable and fairly quick-payback purchase on its own. The marketing treats the thermostat as the star of the show; in most of the homes I actually audit, it’s a supporting player, and treating it as the main fix leads to disappointing results and a mistaken sense that “the smart thermostat didn’t work” when the real issue was never addressed.
Frequently Asked Questions
How much does professional duct sealing typically cost? Depending on your region and duct accessibility, professional duct sealing generally runs $300-1,500 for a full-home job; DIY sealing of accessible joints with mastic sealant can be done for under $200 in materials if the ductwork is reachable in a basement, crawlspace, or attic.
Is it worth getting a professional energy audit before buying a smart thermostat? If your home is more than 25-30 years old and you haven’t had insulation or ductwork updated, yes — many utility companies offer subsidized or free basic energy audits, which is a low-cost way to find out whether a thermostat alone will actually deliver meaningful savings for your specific home.
Do smart thermostats work with older HVAC systems? Most work with standard forced-air systems, but some older systems (certain heat pumps, boilers, or systems without a common “C” wire) need an adapter or aren’t compatible with every smart thermostat model — check compatibility with your specific system before buying, not just the general “works with most systems” claim.
How long does it take a smart thermostat to actually learn a schedule? Manufacturers generally say one to two weeks for the learning algorithm to establish a baseline pattern, though in practice I’ve found it takes closer to a month before the adaptive scheduling stabilizes and stops making noticeable adjustment errors.
Can I get some smart-thermostat savings without buying one, just by manually adjusting my existing thermostat? Yes, to a real degree — manually setting back your thermostat 7-10 degrees for eight hours a day (sleeping or away) can save roughly 10% on heating and cooling costs according to Department of Energy estimates, which captures a meaningful chunk of what a smart thermostat’s scheduling automates for you.
If your home is older, get the ductwork and insulation checked before spending on a smart thermostat — the envelope fixes usually deliver bigger, more reliable savings, and they make whatever thermostat you eventually install actually work as advertised.
