When a smart device stops responding, there are only four places the problem can actually live: the device itself, the radio signal between the device and your router, the router, or the vendor’s cloud service that the device’s app talks to. Each has one question that isolates it. Does a different, unrelated device also fail right now? That points at the router. Does the same device work fine on a completely different network? That rules out the router and points at the radio or the device. Does the manufacturer’s own status page show an outage? That’s the cloud, not your house at all. Working through these four questions in order, rather than guessing, is faster than resetting things one at a time.
Why guessing wastes time
Most troubleshooting advice lists the same four causes side by side (weak signal, outdated firmware, wrong account, router settings) without a way to tell them apart. That leaves you resetting the device, then the router, then reinstalling the app, in whatever order feels right, often repeating steps that already worked. A one-question test per cause gets you to the actual answer faster than a checklist that treats all four as equally likely every time.
Test 1: is it just this device, or everything?
Check whether a second, unrelated smart device also stopped working around the same time. If two or more devices from different manufacturers are failing together, the problem is very unlikely to be inside either device it points at the router or the network itself.
If only one specific device is affected and everything else on the network is fine, move to test 2.
Test 2: does it work on a different network?
Connect the misbehaving device to a different Wi-Fi network entirely a mobile hotspot, or a neighbor’s guest network if you have permission, works for this. If the device connects and behaves normally there, the device itself is not broken. The problem is somewhere in your router or the radio environment around it: a channel that’s become crowded, a band-steering setting pushing it to a radio it can’t use reliably, or a router configuration change.
- If interference or a crowded channel looks likely, see finding the least crowded 2.4 GHz channel.
- If the device seems to drop specifically when it’s close to the router, see band steering and smart home problems.
- If several devices share the same combined network name and only some of them are affected, see splitting your 2.4 GHz and 5 GHz network names.
If the device fails identically on the different network too, the device itself is the more likely cause check its battery (for battery-powered devices), its age, and whether the manufacturer’s app shows a pending firmware update.
Test 3: is the vendor’s own service down?
Some devices show as “connected” in their app, with a green dot or a normal-looking status, but simply won’t respond to commands. This is a different symptom from an outright disconnection, and it usually means the problem isn’t local at all it’s the manufacturer’s cloud service that relays the command from your phone to the device. Check the manufacturer’s own status page or support account (most major smart-home brands publish one) before spending more time on your own network. A widespread outage affecting many customers at once resolves on its own; there’s nothing to fix at your end.
Test 4: none of the above
If a second device isn’t failing, the device behaves the same on a different network, and the vendor’s status page shows everything normal, what’s left is genuinely the device or its immediate radio environment in a way the earlier tests didn’t catch most often an aging battery, a physical obstruction that only affects that one device’s location, or firmware on the device itself that needs updating from within its own app.
Why testing in order beats testing at random
It’s tempting to change several things at once when a device is misbehaving reset it, also update the router, also reinstall the app, all in the same sitting. The problem with that approach is that if the device starts working again, you don’t actually know which change fixed it, which means you can’t tell whether the same thing will happen again next month. Working through the four tests in order, one at a time, tells you exactly which bucket the problem lived in, so you know what to check first if it recurs.
A simple way to keep track
If a device has failed more than once, it’s worth noting three things each time: what the device was doing right before it dropped (nothing, usually (that’s part of the pattern), whether any other device was also affected, and roughly what time it happened. This isn’t about building a spreadsheet) a note in your phone is enough but a pattern across two or three incidents (always around the same time of day, always when a specific other device is also active, always after the router’s weekly automatic restart) points straight at the router or radio environment rather than the device, and saves you from re-running all four tests from scratch every time.
What each branch actually costs to check
None of the four tests requires special equipment, but they do take different amounts of effort:
- Test 1 (another device) costs nothing you either already know another device failed or you don’t.
- Test 2 (different network) costs a few minutes and a phone with a hotspot.
- Test 3 (vendor status page) costs a web search for “[manufacturer name] status page” and a glance.
- Test 4 (the device itself) costs the most checking battery level, physical position, and whether a firmware update is pending inside the device’s own app.
Running them in this order, cheapest first, means you spend the least effort getting to the right answer, rather than jumping straight to the most involved check because it feels like “doing something.”
Building the full picture
These four tests aren’t a one-time exercise they’re the same order worth running any time a previously-working device starts acting up. See a smart device that was working and now drops off for the fuller diagnostic context and why “it worked before” changes where you should look first.
FAQ
What if two devices fail but they’re from the same manufacturer?
That’s weaker evidence than two devices from different manufacturers, because it could still be that specific manufacturer’s cloud service rather than your router. Check the vendor’s status page before concluding it’s local.
Can I run these tests without any technical tools?
Yes. Test 1 just needs you to check another device. Test 2 needs a phone hotspot. Test 3 needs a web browser to check a status page. None of this requires a network analyzer or special software.
What if the device works on a hotspot but the signal there is obviously stronger than at home?
That’s still useful information it tells you the device itself is fine, and the problem is specifically in your home’s radio environment (interference, band steering, or channel congestion) rather than the device being faulty.
How often should I check the vendor’s status page before assuming it’s my network?
Check it first if the device shows as “online” but unresponsive that specific symptom points at the cloud more often than a flat “offline” status does.
Does this order apply to a device that has never connected at all?
No. A device that has never connected is almost always a setup, account, or pairing problem rather than one of these four categories, which assume the device has worked before.