Why smart plugs keep going offline (and the fix order that works)

The symptom: a smart plug that paired fine, worked for weeks, then shows "offline" in the app — sometimes daily, sometimes after a power cut, sometimes for no reason you can name. The plug is rarely broken. The causes have an order, and testing in that order finds the culprit fastest.

Cause 1: the 2.4 GHz requirement (check first)

Most smart plugs are 2.4 GHz-only, and this is the single most common cause. Your router broadcasts both 2.4 and 5 GHz; phones prefer 5 GHz (faster), plugs require 2.4 GHz. Two failure shapes: a band-steering router that moves the plug or its setup traffic onto 5 GHz, or a setup done on a phone that wouldn't drop to 2.4 GHz during pairing. The fixes: give the bands separate names (my-network and my-network-5G) so nothing can be steered, or temporarily disable 5 GHz during pairing. This is also why plugs work fine at the far end of the house where 5 GHz doesn't reach — 2.4 GHz penetrates walls better; the band anatomy is worth knowing.

Cause 2: DHCP address expiry (the weekly-offline signature)

The router loans each device an IP address for a lease period. When the lease expires, a well-behaved plug renews silently; a poorly-implemented one drops offline instead. The tell: the plug disappears on a regular cadence — every few days, same pattern. The fix: in your router's settings, reserve a static IP for the plug (usually listed under DHCP reservations or address reservations). Two minutes of setup, and the failure mode is gone permanently. This single fix resolves a large share of "randomly offline" plugs.

Cause 3: router distance and congestion

2.4 GHz is the crowded band — neighbours' Wi-Fi, microwaves, Bluetooth, other IoT devices. A plug at the edge of range works until congestion peaks, then drops. Diagnose: does it go offline at consistent times (evenings, when every network is busy)? Fix: move the plug one socket closer to the router, or reduce 2.4 GHz congestion — fewer devices on that band, a better router position, or channel selection away from the crowd.

Cause 4: power interruptions (the after-a-storm signature)

A brief outage is a race: when power returns, the router takes 30-90 seconds to boot; the plug reconnects in 5-10 seconds, finds no router, and some plugs give up instead of retrying patiently. The fix is a firmware-behaviour question — update the plug's firmware (makers have progressively improved reconnect logic), and accept that a power-cycle of the plug (10 seconds off, on again) after an outage is normal behaviour, not a defect. A plug on a surge protector also survives the voltage transients that occasionally scramble cheap electronics.

Cause 5: the cloud dependency (the honest limitation)

Many budget plugs route every command through the maker's cloud: your phone → their server → the plug. When their cloud has an outage, your plug is "offline" even though your Wi-Fi is perfect — and this one is not fixable from your side. Diagnose: check the maker's status page or community forums when it happens; if thousands report it simultaneously, it's them, not you. The long-term fix is buying habits: Matter-certified plugs controlled locally don't have this failure mode — local control is one of Matter's actual promises. This is also why a house full of cloud-dependent devices feels flakier than the individual specs suggest.

The fix order, in one line

Confirm 2.4 GHz → reserve the IP → improve signal/congestion → update firmware → check the maker's cloud status. In that order, fastest-test first. And if the plug is offline right now: the 10-second power cycle before anything else — it re-runs the whole connection sequence and clears causes 1, 2 and 4 in one move.

Sources

• Thread Group / CSA — local control and Matter certification