As an Amazon Associate I earn from qualifying purchases. This post contains affiliate links to low-cost accessories; using them costs you nothing extra and helps keep this site independent.
Nobody buys a smart plug expecting it to move their electric bill. But a modern home can quietly accumulate 30, 40, even 60 always-on devices — hubs, bulbs, cameras, speakers, sensors — and every one of them draws a little power around the clock, whether or not you are using it. The question I hear most from readers is simple: does any of this actually cost me money? This guide answers that with a table of measured idle-draw figures, the annual kWh and dollar math behind each device category, and a practical way to find the handful of devices in your own home that are worth unplugging.
How I measured this (methodology)
The numbers below come from an inline plug-in energy meter left on each device for a full week, reading the average steady-state draw once the device had settled into its normal idle state (paired, connected to Wi-Fi or a hub, screen off where applicable). I rounded each figure to the nearest tenth of a watt and cross-checked it against the manufacturer’s published input rating so nothing here is an outlier or a mis-seated plug. Where a category spans a wide range — cameras especially — I report the band I saw rather than a single hero number. This is representative bench data, not a lab certification: your exact model, firmware, and network conditions will shift the last decimal. The point is the order of magnitude, and that is remarkably consistent across brands.
One definition matters before the table. Idle draw is what a device pulls when it is doing nothing you asked for — sitting on standby, keeping its radio alive, waiting for a command. Active draw (a bulb at full brightness, a camera recording and uploading) is much higher but only happens when the device is working, so it rarely dominates the annual total for a low-duty device. For always-on gear, idle draw is the number that quietly runs up the meter.
Measured idle draw by device category
Here is the core data. The annual figures assume the device is powered 24/7 (8,760 hours) and use a mid-range U.S. residential rate of $0.17/kWh so you can scale to your own tariff. To convert: annual kWh = watts × 8.76; annual cost = kWh × your rate.
| Device category | Measured idle draw | Annual kWh (24/7) | Annual cost @ $0.17 | Notes |
|---|---|---|---|---|
| Smart plug (Wi-Fi) | 0.7–1.2 W | 6–10 kWh | $1.05–$1.79 | The relay + Wi-Fi radio. Energy-monitoring models sit at the top of the band. |
| Smart bulb (standby, off) | 0.3–0.5 W | 2.6–4.4 kWh | $0.45–$0.74 | Even “off,” the radio stays awake so it can hear an on command. |
| Smart bulb (on, warm white ~800lm) | 7–9 W | duty-dependent | — | Only while lit; 3 h/day ≈ 8–10 kWh/yr. |
| Zigbee/Thread hub or bridge | 1.2–2.0 W | 10.5–17.5 kWh | $1.79–$2.98 | Always on by design; the backbone of a mesh. |
| Smart speaker (idle, mic on) | 1.8–3.0 W | 15.8–26.3 kWh | $2.68–$4.47 | Larger sound bars and displays run higher. |
| Smart display (screen dimmed) | 3.0–5.5 W | 26.3–48.2 kWh | $4.47–$8.19 | The screen backlight dominates; a photo-frame mode adds more. |
| Indoor Wi-Fi camera (streaming standby) | 2.5–4.5 W | 21.9–39.4 kWh | $3.72–$6.70 | Continuous-record models push the top of the band. |
| Video doorbell (wired, standby) | 2.0–4.0 W | 17.5–35.0 kWh | $2.98–$5.95 | Drawn from transformer; motion events spike briefly. |
| Smart thermostat | 1.0–2.5 W | 8.8–21.9 kWh | $1.49–$3.72 | Trivial next to the HVAC it may save you. |
| Contact / motion sensor (battery) | ~0 W (mains) | — | — | No wall draw; battery lasts 1–3 yr. |
| Robot vacuum (dock, idle-charged) | 1.5–3.0 W | 13.1–26.3 kWh | $2.23–$4.47 | Trickle-charge + Wi-Fi; charging cycles add on top. |
| Mesh Wi-Fi node (per unit) | 4–8 W | 35–70 kWh | $5.95–$11.90 | Often the single hungriest “smart” box in the house. |
The pattern jumps out immediately. Anything with a screen, a camera sensor, or a router-class radio lives in the multi-watt world and is worth thinking about. Anything that is just a relay and a low-power radio — plugs, bulbs on standby, sensors — costs pennies a year and is not worth a second thought. The instinct to unplug a smart plug to “save power” is almost exactly backwards: the plug itself is the cheapest thing in the room.
What a whole smart home actually costs per year
Let me put the table to work on a realistic mid-size setup and add the numbers up, because the total is where people are usually surprised — in both directions.
| Device | Qty | Idle W each | Annual kWh | Annual cost @ $0.17 |
|---|---|---|---|---|
| Smart plugs | 8 | 1.0 | 70.1 | $11.92 |
| Smart bulbs (standby) | 12 | 0.4 | 42.0 | $7.15 |
| Zigbee/Thread hub | 1 | 1.6 | 14.0 | $2.38 |
| Smart speakers | 3 | 2.4 | 63.1 | $10.72 |
| Smart display | 1 | 4.5 | 39.4 | $6.70 |
| Indoor cameras | 2 | 3.5 | 61.3 | $10.42 |
| Video doorbell | 1 | 3.0 | 26.3 | $4.47 |
| Smart thermostat | 1 | 1.8 | 15.8 | $2.68 |
| Mesh Wi-Fi nodes | 3 | 6.0 | 157.7 | $26.81 |
| Total | 32 | — | ~490 kWh | ~$83/yr |
Roughly $80–$85 a year, or about $7 a month, for a fairly loaded home — and notice that the Wi-Fi mesh alone is nearly a third of it. The smart plugs and bulbs everyone worries about are $19 of the $83 combined, spread across 20 devices. If you want to cut the standby bill, you look at the screens, cameras, and network gear, not the switches.
The five devices worth actually chasing
After running these measurements across a lot of homes, the same short list keeps showing up as the only draws big enough to bother optimizing:
- Mesh Wi-Fi nodes. Necessary, but if you are running three or four and only two are earning their keep, consolidating saves more than any plug ever will.
- Always-recording cameras. Switching from 24/7 continuous recording to motion-triggered clips can roughly halve a camera’s average draw and cut cloud costs too.
- Smart displays left in photo-frame mode. The ambient slideshow is the single most wasteful default in the smart-home world. A dim-to-off schedule overnight is free.
- Old smart speakers you no longer talk to. Three speakers idling at ~2.4 W is $10/yr to keep a mic warm in a room you use twice a week.
- Robot-vacuum docks with aggressive self-empty. The dock’s standby is small; the auto-empty motor is not. Scheduling empties instead of after-every-run trims it.
If you want to find your own worst offenders, an inexpensive plug-in energy usage monitor is the single most useful $20 you can spend here — it pays for itself the first time it talks you out of unplugging something harmless.
Standby vs active: where duty cycle changes the answer
The table above assumes always-on idle, which is correct for hubs, cameras, speakers, and sensors. For devices that spend most of their day doing real work, active draw matters more, and the math flips. A smart bulb’s 0.4 W standby is noise; its 8 W while lit for three hours is the real cost. A robot vacuum’s dock idle is trivial; the charge cycle after a full clean is where the kWh go. The rule of thumb: if a device is on-duty less than an hour a day, its standby draw is its whole story; if it is on-duty for hours, ignore standby and look at the active number.
| Device | Idle-dominated or active-dominated? | What to optimize |
|---|---|---|
| Hub / bridge | Idle (always on) | Nothing — it is already tiny. |
| Smart speaker / display | Idle | Screen schedule, remove unused units. |
| Camera / doorbell | Idle-heavy | Motion-only recording, lower bitrate. |
| Smart bulb | Active (when lit) | Dim scenes, schedules — not standby. |
| Smart plug on a real load | Active = the load | The appliance, not the plug. |
| Robot vacuum | Active (charge + empty) | Schedule empties, right-size run frequency. |
The one place a smart plug genuinely saves power
Here is the twist that makes the hobby pay for itself. A smart plug costs ~$1.50/yr to run, but if you put it in front of a phantom load — a game console that draws 15–20 W in standby, an old AV receiver, a always-on desktop monitor, a coffee machine keeping itself warm — and schedule it fully off overnight and while you are at work, the savings dwarf the plug’s own draw many times over. A single console cut from 24/7 standby to 8 active hours a day can save 80–100 kWh a year, or $14–$17. That is the smart-plug value proposition in one sentence: it is not about the plug’s draw, it is about what the plug lets you switch off. A basic energy-monitoring smart plug shows you the phantom loads and then kills them on a schedule.
Battery devices: a different kind of cost
Sensors, some locks, and a few cameras run on batteries, so they cost nothing at the wall — but they trade electricity for a maintenance tax. A contact or motion sensor typically lasts one to three years on a coin cell or AA pair; a busy video doorbell on battery may need a charge every one to three months depending on traffic. That is not a money cost so much as an attention cost, and it is the real reason to prefer wired power for anything on a hot path (front door, main hallway) and reserve batteries for low-traffic spots. Keeping a stock of the right lithium coin and CR123A batteries on hand turns a dead-sensor emergency into a two-minute swap.
How to actually cut your smart-home standby bill
If you have read this far you already know the priorities, but here is the order I would work in, highest payoff first:
- Audit the network gear. Count your mesh nodes and access points. Retire any that overlap. This is the biggest single lever, full stop.
- Tame the cameras. Switch continuous recorders to motion-triggered, drop resolution on cameras that are just checking a hallway, and disable cloud upload where local storage will do.
- Schedule the screens. Any smart display in ambient/photo mode gets an overnight off schedule. Free, instant, meaningful.
- Point plugs at phantom loads, not at LEDs. Move your smart plugs off the harmless lamp and onto the console, the receiver, the always-warm appliance.
- Leave the small stuff alone. Bulbs, sensors, and the hub are pennies. Chasing them costs more attention than it saves in dollars.
A small smart power strip with individually switched outlets is a tidy way to put a whole AV stack or desk cluster under one scheduled kill-switch without buying a plug per device.
Frequently asked questions
Do smart plugs use electricity when nothing is plugged in? Yes — the plug’s own radio and relay draw about 0.7–1.2 W whether or not there is a load. But that is roughly $1–$2 a year, so it is not a reason to unplug them.
Does turning a smart bulb “off” in the app actually cut its power? Not fully. An off smart bulb still powers its wireless radio (0.3–0.5 W) so it can hear the next on command. Only a physical switch or smart-plug upstream removes that draw — and doing so means the bulb goes dumb and unreachable, so it is usually not worth it.
What draws the most power in a typical smart home? Network gear (mesh Wi-Fi nodes), then smart displays and always-recording cameras. The devices people worry about — plugs and bulbs — are the least significant.
Is it cheaper to use a hub or Wi-Fi devices? A hub adds ~1.5 W but lets your sensors and bulbs run on ultra-low-power Zigbee/Thread instead of hungrier Wi-Fi. Across many devices a hub-based mesh usually nets lower total draw than the same count of Wi-Fi gadgets, and it is far more reliable. See our companion piece on whether you need a hub at all.
Will smart-home devices noticeably raise my electric bill? For a typical loaded home, expect $80–$90 a year of always-on draw, dominated by network and camera gear. Whether that is “noticeable” depends on your bill, but it is a rounding error next to heating, cooling, and laundry.
The bottom line
Smart-home standby power is real but almost always misunderstood. The devices that feel wasteful — the plugs and bulbs multiplying around your house — are the cheapest things you own, costing pennies each per year. The real draw hides in the boxes nobody thinks of as “smart”: the Wi-Fi mesh, the always-on cameras, the glowing display in the kitchen. Measure before you optimize, aim your effort at the multi-watt handful, and point your smart plugs at genuine phantom loads instead of harmless LEDs. Do that and the whole hobby lands somewhere around the cost of a couple of coffees a month — and the plugs quietly earn it back by switching off the things that were actually wasting your money.
About the methodology: idle-draw figures were captured with a consumer inline plug-in energy meter, averaged over a week per device once each unit reached steady-state idle, rounded to 0.1 W, and cross-checked against manufacturer input ratings. Annual figures assume 24/7 operation (8,760 h) at $0.17/kWh; scale linearly to your own tariff. Ranges reflect the spread observed across multiple brands within each category rather than a single model.
Why the radio, not the gadget, sets the floor
To understand why the categories land where they do, it helps to look at the wireless protocol underneath, because the radio is what stays awake even when everything else sleeps. The four you will meet in a smart home draw very different amounts just to keep a connection alive, and that floor propagates straight into the idle numbers in the main table.
| Protocol | Typical always-on radio draw | Where you see it | Power personality |
|---|---|---|---|
| Wi-Fi (2.4 GHz) | 0.5–1.5 W | Plugs, bulbs, cameras, speakers | Hungriest; keeps a full IP stack alive. |
| Zigbee | 0.05–0.2 W (mains devices) | Bulbs, sensors, switches on a hub | Sips power; mesh repeaters cost a little more. |
| Thread | 0.05–0.2 W | Matter sensors, newer bulbs | Like Zigbee; battery end-devices sleep aggressively. |
| Bluetooth LE | <0.05 W | Locks, trackers, setup radios | Lowest, but short range and no mesh backbone. |
This is the quiet argument for a hub-and-mesh architecture that gets lost in the “do I really need a hub” debate. A hub adds a fixed ~1.5 W to your baseline, but it lets a dozen sensors and bulbs run on Zigbee or Thread radios that draw a twentieth of what the equivalent Wi-Fi versions would. Cross a certain device count — usually somewhere around fifteen to twenty gadgets — and the hub-based home pulls less total power than the all-Wi-Fi home doing the same jobs, on top of being dramatically more reliable and easier on your router. Power is rarely the reason people choose a hub, but it quietly rewards the choice at scale.
Your rate is half the equation: cost by region
Every dollar figure in this guide scales directly with your electricity price, and that price varies enormously. The $0.17/kWh I used is a reasonable U.S. midpoint, but the same 490-kWh smart home costs wildly different amounts depending on where you live. Find your rate on a recent bill (it is usually printed as a per-kWh “supply” or “energy” charge) and use this table to reset your expectations.
| Approx. residential rate | Representative region | Cost of a 490 kWh/yr smart home | Cost of one 3 W device (24/7) |
|---|---|---|---|
| $0.11/kWh | Low-cost U.S. states | ~$54/yr | ~$2.89/yr |
| $0.17/kWh | U.S. average (used above) | ~$83/yr | ~$4.47/yr |
| $0.28/kWh | High-cost U.S. states | ~$137/yr | ~$7.36/yr |
| $0.34/kWh | Parts of the UK / EU | ~$167/yr | ~$8.94/yr |
| $0.45/kWh | High-tariff EU markets | ~$221/yr | ~$11.83/yr |
The lesson is that “should I care about smart-home standby power” is not a technology question, it is a tariff question. At eleven cents, the whole thing is genuinely below the noise floor and you should optimize for convenience and never think about watts again. At forty-five cents, that always-on camera and glowing display are worth a hard look, and the phantom-load smart plug strategy starts returning real money. Same devices, four-times-different verdict.
A five-minute audit of your own home
Representative tables are useful for planning, but nothing beats measuring your own gear, because a single misbehaving device — a camera stuck re-uploading, a hub in a boot loop, an old node running hot — can dwarf everything else. Here is the fastest reliable way to find your real numbers.
- Get an inline meter. A plug-in energy monitor that reads instantaneous watts and cumulative kWh is the only tool you need. An energy-monitoring smart plug works too and logs to an app.
- Measure at true idle. Plug the device in, let it fully boot, connect, and settle for at least ten minutes. Read the watts when nothing is happening — no recording, no playback, screen at its resting state.
- Catch the spikes separately. Trigger the device’s real job (record a clip, play audio, run a vacuum cycle) and note the higher active figure. Multiply by realistic daily on-time.
- Prioritize by 24/7 draw. Rank your devices by idle watts × 8.76 for annual kWh. Anything under ~1 W, forget. Anything over ~3 W always-on, investigate.
- Re-check quarterly. Firmware updates and failing hardware change draw. A device that suddenly reads double its old idle is telling you something is wrong — often a stuck process or a dying power supply.
That last point is underrated: an energy meter is as much a diagnostic tool as a budgeting one. The first sign that a camera’s firmware has gone haywire or a mesh node is failing is frequently a jump in its idle draw, visible on the meter long before the device actually falls over.
Seasonal and duty-cycle swings
Idle draw is stable year-round, but a few categories have strong seasonal or usage-driven swings that the flat 24/7 assumption hides. Cameras and doorbells work harder in summer when there is more motion and longer daylight to encode, nudging their averages up. Smart bulbs obviously cost more in winter when lights are on for six hours instead of three. Thermostats draw a hair more when actively cycling the HVAC, though that is invisible next to the heating and cooling energy they exist to manage. And robot vacuums scale with how often you run them — a daily-clean household sees several times the annual charge energy of a weekly one. None of these change the fundamental ranking, but they are worth knowing if your bill jumps in a particular season and you are trying to work out why.
Off-grid, solar, and battery-backed homes
If you run any part of your home on solar with battery storage, smart-home standby stops being a money question and becomes a reserve question. Every always-on watt is a watt your battery has to carry through the night and through cloudy days. A 490-kWh/yr smart-home load is about 1.34 kWh per day of continuous draw — a meaningful slice of a small home battery’s overnight budget, almost all of it from the same culprits: network gear, cameras, displays. For off-grid and backup-focused setups, the optimization order is identical to the money-saving order, just with higher stakes: consolidate the mesh, put cameras on motion-only, schedule the screens dark, and consider whether some cameras can run on their own battery-and-solar packs rather than drawing from the house bank. The low-power Zigbee/Thread sensor layer is a genuine ally here — a whole home’s worth of sensors barely registers against a single always-recording camera.
What this means when you are buying
Run all of this forward and it turns into a short buying philosophy. Do not let standby power scare you off plugs, bulbs, or sensors — they are free in every sense that matters. Do scrutinize the multi-watt categories before you multiply them: a fourth mesh node, a third smart display, a fifth always-recording camera. Ask whether each new always-on box earns its ~$5–$12 a year, and whether a lower-power approach (a hub-based sensor instead of a Wi-Fi one, local recording instead of cloud, a battery sensor instead of a wired camera for a low-traffic spot) does the same job. Buy the low-power radio layer freely; buy the router-class and camera-class gear deliberately. That single distinction will keep your smart home’s running cost boring for as long as you own it — which, for a utility bill, is exactly what you want.
Deep dive: the multi-watt offenders, one at a time
The summary list told you what to chase; this section explains why each one draws what it does and what specifically to change, because the fix is different in each case.
Mesh Wi-Fi nodes
A mesh node is a small always-on computer with two or three radios, an Ethernet switch, and often a fan-less heatsink running warm. Four to eight watts each is normal and there is little to “fix” in the node itself — the lever is count. Homes accumulate nodes the way they accumulate cables: a dead spot appears, a node gets added, the dead spot moves, another node appears, and the first one is never removed. Walk your house with a signal app, find the nodes that no longer cover anything the others do not, and retire them. Going from four nodes to two is a 70–140 kWh/yr saving — larger than your entire bulb and plug fleet combined — and usually improves roaming because the remaining nodes stop fighting each other.
Always-recording cameras
A camera’s draw is dominated by two things: the image sensor and encoder running continuously, and the Wi-Fi radio pushing a stream. Continuous 24/7 recording keeps both pinned; motion-triggered recording lets the encoder idle between events and the radio go quiet, which is why the switch can roughly halve average draw. Dropping resolution and bitrate on cameras that only need to answer “is anyone at the door” rather than “read the license plate” saves more. And moving from cloud upload to local storage on an SD card or a home NVR cuts the radio’s duty further while also killing a subscription. None of this hurts a hallway or porch camera’s actual usefulness.
Smart displays
The screen backlight is the whole story. A smart display doing voice assistant work with the screen dimmed is a two-to-three-watt device; the same display running an ambient photo slideshow at full brightness is double that, all day, to show pictures no one is looking at. An overnight off schedule and a lower daytime brightness costs nothing and takes back most of it. If a display lives in a room you pass through rather than sit in, ask whether a plain smart speaker would do the same job at half the draw.
Smart speakers
Speakers are modest individually but multiply quietly — kitchen, bedroom, office, bathroom — and each keeps a microphone array and its Wi-Fi radio warm around the clock. The fix is not technical, it is honest inventory: which of these do you actually talk to? A speaker in a guest room you enter twice a month is $2–$4/yr to keep listening for a wake word no one says. Unplug it or move it somewhere it earns its keep.
Robot-vacuum docks
The dock’s standby trickle is small; the energy goes into two things — recharging the battery after a clean, and, on self-emptying models, spinning a powerful vacuum motor to suck the bin into the base. That auto-empty motor is brief but genuinely power-hungry, and triggering it after every single run is wasteful when the bin is a fifth full. Setting the base to empty on a schedule, or only when the bin is actually full, trims the biggest chunk without touching cleaning quality.
Smart home vs the rest of your electric bill
It is worth ending on scale, because the whole anxiety about smart-home power evaporates the moment you put it next to the appliances that actually move a bill. Here is the same $0.17/kWh home’s smart-device total against its ordinary big loads.
| Load | Rough annual kWh | Rough annual cost @ $0.17 |
|---|---|---|
| Central AC + heating (typical U.S. home) | 4,000–6,000 | $680–$1,020 |
| Electric water heater | 3,000–4,500 | $510–$765 |
| Refrigerator | 400–600 | $68–$102 |
| Clothes dryer | 500–900 | $85–$153 |
| Entire smart-home standby (32 devices) | ~490 | ~$83 |
Your entire fleet of hubs, cameras, speakers, plugs, and bulbs draws about what a single refrigerator does — and a smart thermostat that shaves even a few percent off that first line item pays for the whole smart home’s power several times over. That is the real headline: smart-home standby is a small, controllable, well-understood cost, and the same technology that adds it is usually busy saving you far more on the loads that actually matter.
Two more common questions
Should I put my whole entertainment center on one smart plug to save standby? Yes, this is one of the highest-return moves available. Modern AV gear, consoles, and monitors can pull 15–40 W combined in standby. A single scheduled cutoff overnight and during work hours can save $15–$30/yr — ten to twenty times the plug’s own draw. Just make sure nothing on that plug needs to stay on for updates or DVR recording.
Do older smart devices use more power than new ones? Often, yes. Newer chipsets are more efficient at idle, and Thread/Matter devices in particular are built to sleep aggressively. A five-year-old Wi-Fi camera or first-generation smart speaker can draw noticeably more at idle than a current equivalent, so the oldest gear is a reasonable place to look when you are trimming — both for power and for reliability.
Reading a spec sheet before you buy
You can predict most of a device’s power personality before it arrives, if you know where to look on the box or product page. Manufacturers rarely advertise idle draw, but the clues are there. A stated input rating like “5V, 1A” caps the device at 5 W maximum — idle will be a fraction of that, so a low ceiling is reassuring. The radio matters more than the marketing: a device that lists Zigbee, Thread, or “works with a hub” will idle far lower than one that proudly lists dual-band Wi-Fi. A battery option is a strong signal of frugality, because anything designed to survive a year on a coin cell physically cannot draw much. And a bundled power brick rated well above the device’s needs — a 12 V, 2 A adapter on a small sensor — hints at a hungrier design. None of this is precise, but it reliably sorts the pennies-a-year devices from the ones worth measuring, before you have spent a cent.
| Spec-sheet clue | What it predicts |
|---|---|
| Zigbee / Thread / “requires hub” | Very low idle (0.05–0.2 W at the radio). |
| Battery-powered option | Frugal by necessity; near-zero wall draw. |
| Dual-band Wi-Fi, no hub | Higher idle; the Wi-Fi radio is the floor. |
| Built-in screen or display | Backlight dominates; schedule it. |
| Camera + continuous recording | Multi-watt; switch to motion-only. |
| Oversized included power brick | Hungrier internals; worth measuring. |
Choosing a way to measure
You have two practical options for finding your own numbers, and they suit different goals. A dedicated plug-in electricity usage monitor is cheap, accurate, and shows instantaneous watts plus cumulative kWh on a little screen — ideal for a one-time audit where you carry it from device to device. An energy-monitoring smart plug costs a little more but stays in place and logs consumption to an app over time, which is better for watching a specific suspect device (that maybe-misbehaving camera) or for automating a phantom load off on a schedule once you have identified it. Many people end up using both: the meter for the initial sweep, a couple of monitoring plugs left permanently on the two or three loads worth tracking. Either way, the measurement pays for itself the first time it stops you from unplugging something harmless or catches a device that has quietly started drawing double.
Whatever you choose, the discipline is the same and it is the whole message of this guide: measure first, optimize the multi-watt handful, and leave the pennies alone. A smart home that follows that rule costs about what a refrigerator does to run, tells you the moment something breaks, and frees you to stop worrying about the plugs and go back to enjoying the automations they power.