Status, 1 November 2026: planned. No measurements have been taken yet. This page sets out exactly how Look Into will run its first hands-on test, before any result exists, so the method cannot be bent to fit the findings. When results are published, they will link back here, and any change to this protocol will be dated in the change log below.
The question
How much power do ten everyday household devices draw when they are left plugged in and not being used, and what does that cost a year? Our UK Always-On Load Index prices these loads against the legal standby limits, because almost no product says what it actually draws. This test measures ten of them.
Who measures, and where
Jacob Whitmore, Look Into's founder, takes every reading at his home in the UK, on devices the household already owns or has bought at full retail price. No device is supplied by a maker or retailer. Each device is identified on the results page by brand, model and age, so readers know which unit was measured.
The ten devices
- Broadband router (always on)
- Television in standby
- Set-top or streaming box in standby
- Smart speaker, idle
- Games console in rest or standby mode
- Microwave with clock display, idle
- Phone charger plugged in with no phone attached
- Laptop charger plugged in with no laptop attached
- Electric toothbrush on its charging base, fully charged
- Printer, idle
If the household does not own one of these, the results page will say so and the device will not be replaced by a different one.
The meter
A plug-in energy meter with a UK 13A plug whose maker states its accuracy, for example ±2% or better. The results page will name the exact model, its stated accuracy and resolution, and link to the maker's specification. A meter's stated accuracy usually applies to its normal range; very small loads, below about 2 watts, may sit outside it. We treat those readings with extra care (see below) and say so next to each one.
How each device is measured
- Settle. Plug the device into the meter in the state being tested (standby, idle or charger only) and leave it for 30 minutes before recording anything.
- Spot readings. Record the watts shown three times, at least ten minutes apart, with the time of each reading.
- Energy over time. Reset the meter's energy counter and leave the device on it for 24 hours, or 72 hours for any device whose spot readings are below 2 watts, then record the kWh. Average watts = kWh × 1,000 ÷ hours. This is the figure the cost is based on, because it captures devices that wake up and draw more from time to time.
- Repeat. Run the energy measurement a second time on a different day. If the two averages differ by more than 20%, run a third and report all three.
- Record. A photograph of the meter display at each energy reading, with the device visible where practical. Photographs are published with the results.
How the cost is worked out
Yearly cost = average watts × 8,760 hours ÷ 1,000 × the electricity unit rate under the Ofgem price cap in force on the day of publication. At the 1 October to 31 December 2026 average of 26.32p per kWh, one watt left on all year costs about £2.31 (8.76 kWh × 26.32p). The results page will state the rate it uses and show the arithmetic for every device.
What will be published
- Every reading, including any that look odd, with its date and time, in a table and a downloadable CSV.
- The meter model and its stated specification.
- Each device's brand, model, approximate age and settings (for example, "instant on" or "eco" mode).
- The photographs.
- A comparison with the maker's stated standby figure, where the maker publishes one.
What this test cannot tell you
- It measures one unit of each device, in one home. Another unit, firmware version or setting may draw more or less.
- A plug-in meter is not a laboratory instrument. Readings below about 2 watts are the least certain, and are marked as such.
- It does not test safety, performance or how long a device lasts.
Independence
No maker or retailer has seen this protocol or will see results before publication. The results page will carry no affiliate links. It follows our investigation standards, and once published its figures will carry the label "Look Into testing".
Change log
- 1 November 2026: protocol published before any measurement was taken.