How We Test Smart Home Devices

The Test Environment

There's one test environment here, and it's a house being built and wired room by room. Home Assistant runs the logic on local hardware in the utility room. Google Home is the layer my family actually uses, so anything that only works from a vendor app has already half failed. Somfy motors drive the blinds and external shades through TaHoma, exposed back into Home Assistant. New hardware gets installed into that stack, not onto a bench next to it.

That matters more than it sounds. A sensor that pairs perfectly on a clean network behaves differently in a house with 60-odd Zigbee devices, thick walls, and an unfinished floor full of dust. The problems I write about are the ones that showed up in the fabric of a real building.

What Goes Into a Review

Not every product on this site passes through our hands, and we won't pretend otherwise. The catalogue here is far larger than one house, and a review draws on a mix of sources depending on what we could get access to. Every review is built from some combination of the following, and we try to be clear on each page about which applies.

• Hands-on testing of units we were able to obtain, installed in the house and left running for days or weeks. Where a review is based on direct testing, it says so with specifics: firmware versions, wiring notes, and problems we actually hit.
• Manufacturer specifications and datasheets, cross-checked against listed compatibility rather than taken at face value. When we haven't tested a unit ourselves, our assessment leans on documented specs and we frame it as such.
• Verified owner feedback aggregated from retailer reviews, community forums, and support threads, used to surface recurring reliability issues that a short test window can miss.

What We Check

For devices we test directly, we look at setup difficulty, app quality, and how reliably the product integrates with the three systems this house runs on. A device gets checked for local control in Home Assistant, clean exposure to Google Home, and where it's relevant, whether it cooperates with the Somfy side of the house. Compatibility claims on the box get verified one at a time rather than taken on trust. We also note firmware update behaviour and any stability issues that show up over the test period, plus anything that would have been easier to deal with before the walls were closed.

For products we assess from specifications and owner feedback, we focus on measurable, documented differences: sensor ranges, battery figures, protocol support, and the compatibility claims that most often trip people up. We flag where a spec is manufacturer-stated rather than independently confirmed.

Honesty About Limits

A two-week test tells you about setup and early reliability, not about how a device holds up over years. Owner feedback helps fill that gap, but it isn't a controlled test. One house is also one house: my walls, my router, my climate. A blind motor that copes here might struggle on a wider window or a colder wall. Where we're uncertain, we say so, and we'd rather flag a limitation than paper over it. If we get something wrong, tell us and we'll fix it.

Affiliate Disclosure

Some links on this site are affiliate links, which means we may earn a commission if you buy through them at no extra cost to you. That never changes a rating or which products we recommend. Full details are on our affiliate disclosure page.