A plug that reports watts is genuinely useful. It is also not a meter, and treating the number as if it were one produces confidently wrong conclusions about what appliances cost to run.
How the measurement is made
Inside the plug is a small metering chip measuring current through a shunt or a current transformer, and voltage across the mains. Multiplying the two gives apparent power in volt-amperes. Real power in watts is apparent power times the power factor, and the metering chip estimates power factor from the phase relationship between the two waveforms.
That estimate is good for a resistive load, where current and voltage are in phase: a heater, a kettle, an incandescent lamp. It is much weaker for a switch-mode supply, which draws current in sharp pulses at the peak of the voltage waveform. A phone charger, an LED driver, a laptop brick, a television: all of these have a distorted current waveform and a power factor well below one, and the cheaper the metering front end, the more the reported watts drift from the truth.
Where the error shows up
The worst case is small loads. A shunt sized for a 16 A plug has most of its resolution spent at the top of the range, so a 5 W standby reading can be off by a large fraction and can even read zero. That is why a plug will confidently tell you an idle television draws nothing when it draws several watts, and why summing standby readings across a house is the least reliable thing you can do with this data.
Inductive loads are the second case. A motor, a pump, a fridge compressor, draws current that lags the voltage and a startup surge many times the running figure. The running number will usually be plausible, but the surge happens in a fraction of a second and most plugs sample too slowly to catch it, so the one event that matters for sizing a circuit is the one the plug does not record.
What it is good for
Trends, not absolutes. If the same plug reports 90 W yesterday and 140 W today for the same appliance, something changed and that is worth investigating. It is good for detecting that a device is on at all, for spotting a heater that cycles more often than it should, and for comparing two appliances of the same type on the same plug, because the same error applies to both readings.
It is not good for billing, for verifying an electricity supplier, or for adding up a whole house. For those you want a revenue-grade meter at the supply, which is calibrated and measures energy rather than reporting instantaneous power to a phone app.
Practical rules
Treat anything under about 10 W as indicative only. Do not compare readings between two different plug models, because the front ends differ more than people expect. And if you need to know whether a device is worth replacing for running cost, measure it over a week and use the energy total, not the instantaneous figure you happened to glance at when the compressor kicked in.
Where this matters when you are buying
Check the hardware first: smart switches.
Panel standards, protocols and MOQ live on the Xingliao smart home factory site; panel families (86, 118, 147) are on the smart switch board manufacturing page. Sourcing questions go to the factory team.
Top comments (0)