Home energy monitors: what they measure and what they cannot

How this page is funded: No affiliate relationship is in place. Nothing on this page earns this site a commission. Every figure here is computed from the published data in Sources, by the method shown in full. More

Three kinds, and they answer different questions

They are usually sold as one category. They are not one category, and buying the wrong type is the most common way people end up disappointed.

Prices are typical street prices, not a quote. Installation in the panel is an electrician’s job in most places.
Type	Where it goes	Answers	Rough cost
Plug-in meter	Between a socket and one appliance	Exactly what this one device uses	$20–40
Whole-home clamp	On the mains in your panel	Total draw, live, and what turned on just now	$120–200
Circuit-level	Clamps on individual breakers	Which circuit, all the time, without guessing	$200–450

What a whole-home monitor genuinely gives you

A live total, at a resolution utilities do not provide. Your utility shows you a month; a monitor shows you the second the water heater fired. That granularity is what turns “the bill is high” into “the backup heat came on nine times last night”.

The second thing is a baseline. Watch the total at three in the morning when nothing should be running. Whatever that number is, it is your always-on load, and it is running 8,760 hours a year.

What they cannot do

Identify appliances reliably by signature alone. Machine-learning disaggregation — “we detected your dishwasher” — is genuinely impressive on large, distinctive loads and unreliable on everything else. Treat named appliances as a hint, and circuit clamps as the answer.
Save you anything on their own. A monitor is an instrument. It changes the bill only if it changes what you do, and the change is nearly always a schedule, a setpoint or a device you retire.
Survive a panel with no room. Check you have physical space around the breakers and that the clamps fit your conductor size before ordering. This is the most common reason one gets returned.

When the cheap option is the right one

If you have one suspect — a garage refrigerator, a dehumidifier, a hot tub pump — a plug-in meter answers the question for about the cost of $30 and two weeks of patience. A whole-home monitor is the right buy when you do not have a suspect, or when the load you care about is hard-wired and has no plug to meter: a water heater, a well pump, baseboard heaters, an EV charger.

How to choose

Decide whether you have a suspect. One suspect means a plug meter. No suspect means whole-home.
Check the panel physically — space around the breakers, conductor diameter, and whether your service is single or split phase.
Prefer a monitor that exports raw data. Anything that only shows you its own app will eventually stop showing you anything.
Ignore the appliance-detection marketing. Buy on circuit count and data access, then treat detection as a bonus.

Where to look

Two manufacturers dominate the whole-home category and they solve the problem differently. WattDraw has not tested either — this site does not publish reviews of hardware nobody here has measured — so what follows is what each one measures, which is the part you can check against your own question before spending anything.

Both need panel access. Neither is a substitute for a plug meter when you already have one suspect.
Approach	How it identifies a load	Best when
Emporia	Physical clamps on individual breakers — up to 16 circuits. It knows which circuit because it is measuring that circuit.	You want an answer rather than an inference, and you can get an electrician into the panel.
Sense	Two clamps on the mains plus machine learning that infers appliances from their electrical signature.	You want whole-home totals and live feedback, and treat named appliances as a hint.

What to check before you order

Five things, in the order they cause returns. The first three are physical and no amount of software fixes them.

Space in the panel. Each circuit clamp needs room around its breaker and a route back to the sensor unit. A tightly packed 40-slot panel with a full complement of AFCI breakers may simply not have it. Photograph the inside of your panel cover before ordering and compare it with the manufacturer’s installation photos.
Conductor diameter. Mains clamps are sized to the service conductors. A 200 A service uses cable a 100 A clamp will not close around. Check the clamp’s stated maximum against your service size.
Service type. North American residential is normally split-phase 240 V with two hot legs, and a monitor must measure both or it reports roughly half your consumption. Any monitor sold for the US market handles this — but check if you are buying from elsewhere.
Where the sensor unit sits. It needs power inside the panel and a Wi-Fi signal there. Panels are often in basements, garages and metal enclosures, all of which attenuate Wi-Fi badly. A monitor that cannot reach the network logs nothing.
Whether the data leaves the app. Look for a documented API, a CSV export, or local access such as MQTT. A device whose readings exist only inside its manufacturer’s app stops being useful the day that app changes, and you cannot check its arithmetic against your own bill.

What the readings are actually worth

Three questions a monitor answers well, and one it does not.

The first three are why a monitor earns its money. The fourth is what the marketing leads with.
Question	How well a monitor answers it
What is my always-on load?	Perfectly, on the first night. Read the total at 3am when nothing should be running; that figure runs 8,760 hours a year.
Did the backup heat come on?	Very well. Resistance strip heat is a large, unmistakable step change in draw, and seeing it is often worth the price of the device on its own.
Which circuit is expensive?	Exactly, if you bought circuit clamps. Not at all, if you bought whole-home only.
Which appliance is that?	Unreliably, outside large distinctive loads. Treat a named appliance as a hypothesis to test with a plug meter, not as a measurement.

At 18.34¢/kWh. Subtract your refrigerator — around 48 W average — before judging the rest.
If your 3am reading is…	Over a year	Costs	Verdict
30 W	263 kWh	$48	Lean. Nothing to find here.
60 W	526 kWh	$96	Normal for a connected household.
100 W	876 kWh	$161	Worth an hour with the breakers.
150 W	1,314 kWh	$241	Something substantial is running. Start switching circuits off one at a time.
250 W	2,190 kWh	$402	Something substantial is running. Start switching circuits off one at a time.

Questions

Will an energy monitor lower my bill?

Only by telling you something you then act on. The realistic mechanism is that it identifies one or two loads you were not counting — an always-on device, a heater engaging more than expected, an appliance running longer than it should — and you deal with those. If you already know where your electricity goes, it will confirm it and save you nothing.

Do I need an electrician?

For anything that clamps inside the panel, yes in most jurisdictions, and yes for your own safety regardless. Plug-in meters need nobody. Some monitors offer a version that sits outside the panel on the meter itself, which avoids the issue entirely where your meter supports it.

Is my utility’s own app good enough?

Increasingly, often yes. Many smart meters now report hourly or even 15-minute interval data through the utility portal, free, and hourly data is enough to spot a heater cycling overnight or a pump running when nobody is home. Check what your utility already gives you before spending anything.

All guides · How the figures are computed · Break down your own bill