It always arrives the same way: an email with a subject line like “Important changes to your [device] service.” Sometimes it’s a clean shutdown with a date. Sometimes the app just starts throwing errors and the support forum fills up. Either way, a thing you bought — a camera, a thermostat, a hub, a lock, a doorbell — is about to stop doing what it did, because the company that made it has decided the servers it depends on aren’t worth running anymore.
This is now a routine event, not a freak one. Insteon went dark overnight. Wink put a gun to its users’ heads with a surprise subscription. Iris by Lowe’s, Staples Connect, Revolv (bought by Google, then bricked on purpose), Best Buy’s Insignia line, Chamberlain locking out third-party MyQ access — the list is long and gets longer. This post is the survival guide: what to do the day the email lands, what to do before it lands, and how to evaluate a vendor’s longevity before you spend money. It pairs closely with our piece on what happens when a smart-home vendor folds — that one is the autopsy; this one is triage.
Step 1: Figure out what actually died
“Loses cloud support” covers a range of outcomes, and the right response depends on which one you got:
- The whole product bricks. No app, no local control, dead hardware. Worst case (see: Revolv). Nothing to salvage; skip to replacement.
- The app dies but the device keeps doing its last job. A thermostat holds its schedule, a switch keeps switching on its timer, but you can’t change anything remotely. Stable for now, but a ticking clock — don’t reset it, because it may not come back.
- Cloud features go behind a paywall. The device works; history, alerts, or remote access now cost a monthly fee. This is the subscription-creep play, and you get to decide whether to pay the ransom or migrate.
- Integrations break but local stuff works. The device still functions on its own network or hub, but the HomeKit/Alexa/Google bridge is gone, or the open API got locked down. Often the most recoverable case — a local controller can frequently pick it back up.
Before you touch anything, write down the firmware version, export any data the app still lets you export (camera footage especially), and screenshot your configuration. A factory reset on a cloud-dependent device after the cloud is gone often turns a “degraded but working” device into a paperweight.
Step 2: Check whether Home Assistant can rescue it
The single most useful move when a device loses its cloud is to put a local controller in front of it. Home Assistant in particular has integrations — official and community — for a startling number of products whose manufacturers have walked away, because the local protocol (Zigbee, Z-Wave, Matter, a LAN API, even reverse-engineered cloud endpoints) often keeps working long after the official app dies.
Zigbee and Z-Wave devices are the easy win: pair them to a generic coordinator and a Home Assistant box and they don’t care that the original hub company is gone. Wi-Fi devices are hit-or-miss — some have a local API, some can be flashed with open-source firmware (ESPHome, Tasmota) if they’re built on common chips, some are locked down hard. Our Home Assistant vs HomeKit vs Alexa comparison covers why a local-first hub is the resilient choice in the first place; the orphaned-device rescue scenario is exactly the case it’s built for. If the device is still functional today, the move is to migrate it onto a local controller now, while it still works, rather than after it’s dead.
Step 3: Decide migrate-or-replace, by category
Some categories are worth the effort to rescue; some you should just replace with something that doesn’t have this failure mode. Roughly:
Cameras & doorbells — usually replace
Cloud cameras that lose their cloud are mostly unrecoverable, and even the ones with an RTSP stream are a pain to live with. This is the category where “it’s on someone else’s server” hurts most. The right answer is to move to a local NVR platform — the local NVR vs cloud cameras and UniFi Protect vs Ring vs Nest posts lay out the case. If your old setup was a Ring or similar that “works but now costs a subscription,” the doorbell post explains why a wired, on-prem doorbell is the upgrade, not just a swap.
Locks & access control — depends on the guts
A smart lock that loses cloud but keeps Bluetooth, a keypad, and a physical key is annoying, not dangerous — you’ve lost remote lock/unlock and audit logs. A Matter or Z-Wave lock can usually be re-homed to a local controller. A lock that was fully cloud-dependent for everything, or a proprietary commercial system whose vendor folded, should be replaced — for any business or rental, that’s urgent, and our access control for small offices post covers on-prem controllers that don’t have this exposure. Don’t wait on this one; a door that won’t lock is a different class of problem than a thermostat that won’t schedule.
Thermostats & HVAC — often rescuable
Most thermostats keep running their last schedule even with no cloud, and many speak a local protocol or have a documented API. A Z-Wave or Matter thermostat moves to Home Assistant cleanly. Even some cloud-first thermostats have local fallback. If you have to replace one, pick a model with local control and heat-pump-aware logic — our posts on why smart thermostats don’t save Utah homes money and heat-pump-aware automation cover what to look for.
Lighting & switches — almost always rescuable
In-wall smart switches and dimmers on Zigbee, Z-Wave, Lutron’s Clear Connect, or Matter are about the most durable thing in the smart home — the radio protocol outlives the app. If the manufacturer’s cloud dies, a local controller picks them up. This is one more reason the switches-vs-bulbs post comes down where it does: switches on a real protocol are infrastructure; cloud-only bulbs are disposable.
Hubs & bridges — the painful one
When the hub is what dies (Wink, Iris, Insteon), everything paired to it goes with it. The recovery is to stand up a new local hub — Home Assistant with Zigbee/Z-Wave radios, or a Hubitat — and re-pair the devices that speak open protocols. Devices that only ever spoke to the dead hub’s proprietary cloud are lost. This is why we push people toward a vendor-neutral hub from the start.
Step 4: Don’t let it happen again — vetting vendor longevity
Before you buy the replacement, run the new candidate through the same filter we’d use:
- Does it work fully without the internet? If pulling the WAN cable kills core function, you’re renting, not owning. On-prem recording, local API, local automation — these are the marks of something that survives its maker.
- Does it speak an open protocol? Matter, Zigbee, Z-Wave, ONVIF, RTSP, a documented LAN API. Open protocol means a third party can keep it alive. Proprietary-cloud-only means you’re one business decision from a brick.
- Is the company’s incentive aligned with keeping it running? A vendor that sells hardware at a healthy margin and offers an optional cloud service has reason to keep the lights on. One that sold the hardware at a loss to hook you on a subscription — or got acquired by a giant with a history of killing products — does not.
- What’s the published support window? Real infrastructure vendors state end-of-life and firmware-support timelines. Silence is an answer.
- Is there an exit? If the company vanished tomorrow, could you keep using the device locally, or export your data, or flash open firmware? If the honest answer is “no,” price that in.
That checklist is, not coincidentally, why the systems we install — UniFi Protect recording to a drive in the rack, on-prem access controllers, switches and sensors on standard radio protocols, automation on a local hub — don’t generate the “important changes to your service” email. It’s also the difference the $500 vs $5,000 security system post is really about: the cheap one is cheap partly because someone else owns the part that can be switched off.
Bottom line
When a smart-home device loses cloud support: stop, document, and export before you reset anything; check whether a local controller like Home Assistant can adopt it; and decide migrate-or-replace by category — cameras and proprietary hubs usually get replaced, switches and most thermostats usually get rescued, locks depend on the guts. Then close the loop by buying a replacement that works offline, speaks an open protocol, and comes from a vendor whose business model doesn’t depend on holding your house hostage. The cloud is a convenience. It should never be the thing keeping your home running.
Keystone Integration migrates orphaned smart-home gear to local, customer-owned platforms across Holladay, Sandy, Draper, Park City, and the rest of the Wasatch Front — on-prem recording, local control, no cloud fees, nothing that bricks when a vendor walks away. See the full service list or get in touch and we’ll triage what you’ve got.