In April 2022, Insteon — a company that had been selling smart-home hubs, switches, and dimmers since 2005 — disappeared. No email, no announcement, no transition plan. One day the cloud servers went offline, and every Insteon device in every home that depended on the Insteon Hub stopped responding. Lights, locks, scenes, schedules — all gone. Seventeen years of hardware, bricked by a server shutdown.
In 2023, Wink — once a Best-Buy-featured smart-home hub — quietly went subscription-only, then stopped shipping updates, then went functionally dark. In 2024, the original Samsung SmartThings hub was deprecated. In 2025, multiple smaller IoT platforms (Wyze’s free cloud tier, Chamberlain’s open API) were either paywalled or shut down.
This isn’t a prediction about the future. It’s the pattern that’s already happened, repeatedly. If your smart home depends entirely on a single vendor’s cloud, that vendor’s business model is your single point of failure.
Why this keeps happening
Smart-home companies, especially the VC-funded ones, have a fundamental business-model tension: the hardware is sold at or near cost to acquire customers, and the plan is to monetize later through subscriptions, data, or ecosystem lock-in. When the subscription revenue doesn’t materialize — because consumers hate subscriptions for things that used to be free — the company either pivots to a paywall (Wink, Ring raising prices, Nest Aware price hikes), gets acquired and deprioritized (Insteon, Revolv under Nest/Google), or simply shuts down.
The homeowner is always the last to find out. And by then, the gear they bought is either useless or degraded.
The subscription creep problem
Even vendors that don’t disappear have a clear pattern: launch with a free tier to build market share, then gradually move features behind a paywall. We covered this in detail with Ring and Nest’s camera subscription models: Ring Home launched at $3/month and is now $4.99– $19.99; Nest Aware went from $6 to $8 to $15 to $20/month over its lifetime. Chamberlain added a MyQ subscription for features that were free at launch.
None of these are going away, but the trajectory is clear: every feature you rely on that runs through someone else’s cloud is a feature that can be paywalled later.
What “customer-owned” actually means
This is the core principle behind how we design systems: the homeowner should own the hardware, own the data, and be able to keep things running even if a vendor pivots, gets acquired, or dies. In practice that means:
- Local processing, not cloud processing. The brain of the system runs on hardware in your house, not on a server that can be turned off.
- Local storage for cameras. Your footage lives on a hard drive in your network closet, not in someone else’s S3 bucket.
- Standards-based communication. Devices that speak Zigbee, Z-Wave, Thread, Matter, or plain Wi-Fi to a local controller can survive a vendor change. Devices that only speak a proprietary cloud protocol (old Tuya, old Insteon) cannot.
- Replaceable vendor, not replaceable house. If Sonos goes dark tomorrow, the in-wall speakers and the Cat6 runs to the rack are still there — you swap the Amp for a different streaming amplifier. If UniFi goes dark, the cameras are PoE on Cat6 — you swap the NVR and the cameras work on any ONVIF-compatible system. The infrastructure outlives the brand.
The platforms that survive a vendor exit
Home Assistant
Open-source, runs on a $100 box in your closet, integrates with essentially every smart device protocol. If any single integration breaks, the community usually patches it within weeks. Your automations, your data, your hardware. It requires more setup than an off-the-shelf hub, but it is the most vendor-proof platform that exists in 2026.
UniFi (for networking, cameras, and access control)
The controller, the NVR, and all processing run on a device you own in your house. Ubiquiti is a publicly-traded company with a large install base, which is about as safe as any vendor bet gets. But even if it weren’t — the cameras are standard PoE, the switches are standard Ethernet, the cables are Cat6. Swap the controller and the infrastructure still works.
Matter / Thread
The new cross-vendor smart-home standard (backed by Apple, Google, Amazon, Samsung, and others) is specifically designed to decouple devices from vendor clouds. A Matter light switch works with any Matter controller. It’s still early — not everything supports it yet — but it’s the first credible attempt at a vendor-neutral device protocol that has real industry weight behind it.
What to look for when you’re choosing gear
You don’t have to go full self-hosted to be resilient. A few questions that filter out the worst vendor-dependency risks:
- Does it work without the internet? If your Wi-Fi stays up but your ISP goes down, do the lights still respond? Do automations still fire? Do cameras still record? If the answer is no to any of these, the cloud is load-bearing and you’re exposed.
- Is there a local API or a standard protocol? Z-Wave, Zigbee, Thread, Matter, ONVIF (for cameras), MQTT — these all mean you can swap the controller without replacing the devices.
- What happens if I stop paying? For subscription-gated products, what exactly stops working? With Ring: no cloud recording. With Nest: 3-hour history only. With UniFi Protect: nothing stops working. That’s a real differentiator.
- How old is the company, and how do they make money? A company selling hardware at a profit has a sustainable model. A company selling hardware at a loss to get you into a subscription funnel has a model that works until it doesn’t.
The infrastructure layer is what survives
This is the lens we use on every install: separate the infrastructure from the vendor layer. The Cat6 in the walls, the conduit to the gate, the speaker wire to the ceiling, the rack in the closet — none of that cares about firmware versions, app rewrites, or subscription changes. It’s copper and fiber. It works in 2026 and it’ll work in 2036.
The vendor layer — the specific brand of camera, the specific smart-lock model, the specific app — is the part that changes. Design the house so the vendor layer is swappable, and you’re protected against the next Insteon, the next Wink, and the next subscription hike.
Bottom line
Smart-home vendors will continue to raise prices, deprecate products, and occasionally collapse. The way to build against that is to own the infrastructure, choose standards-based protocols, prefer local processing over cloud processing, and design the system so the brand on the box is the most replaceable part of the stack.
Keystone Integration designs vendor-resilient smart-home systems across Alpine and the rest of the Wasatch Front — customer-owned gear, local processing, standard cabling, and systems that outlive the brand on the box. You can see the full list of what we do on our main site, or get in touch to scope a system built to last.