Do You Actually Need a Smart Home Hub in 2026?

For years the answer was easy. Yes, you need a hub, because nothing worked with anything else. Then Matter arrived, promised to fix the mess, and a lot of people concluded that hubs were finished.

They aren't, but the reason to own one has changed. It's no longer about compatibility. It's about what happens when things go wrong, and what happens when you own a lot of devices rather than a few.

First, check whether you already own one

This is the step most people skip. A hub is not necessarily a separate box you go out and buy.

Most recent smart speakers and displays quietly include hub hardware. The Amazon Echo Hub has Zigbee, Thread, and Matter radios built in. The HomePod mini is a Thread border router and a Home hub. The Google Nest Hub is a Matter controller.

The Apple TV 4K needs a warning, because this is where people get caught. Only the 128 GB Wi-Fi + Ethernet version has a Thread radio. Apple's own spec sheet lists Thread under that model and no other, so the cheaper 64 GB Wi-Fi box is a Home hub and a Matter controller over Wi-Fi, and it is not a Thread border router. If you are buying an Apple TV specifically so your Thread sensors have something to talk to, buy the right one.

So before you spend anything, find out what the device already sitting on your kitchen counter actually does. A surprising number of people buy a hub to solve a problem that the speaker they already own would have solved.

When you genuinely don't need one

Be honest about your setup. If all of these are true, skip the hub and spend the money on devices instead.

You have a small number of devices. A few Wi-Fi bulbs and a couple of smart plugs talk to your router directly and work fine. Adding a hub to that adds a box, not a capability.

Everything you own is Wi-Fi. If nothing in your house speaks Zigbee, Thread, or Z-Wave, one of a hub's main jobs simply doesn't apply to you.

You don't mind cloud dependence. If you can live with voice control going quiet during an internet outage, that removes the other main argument.

You aren't planning to expand. The case for a hub grows with the size of the system. At six devices it's marginal. At forty it isn't.

When a hub genuinely earns its place

You want the house to work when the internet doesn't. This is the strongest argument, and the one people only appreciate after their first outage. A hub can run automations locally: lights still respond to sensors, routines still fire, and your home doesn't go dumb because a data centre had a bad afternoon. Most cloud-first Wi-Fi devices stop accepting commands entirely in an outage, even though your phone and the bulb are sitting on the same network, two metres apart.

You have a lot of devices. Wi-Fi was not designed for forty chatty low-traffic gadgets, and consumer routers get unhappy. Zigbee and Thread were designed for exactly that, and they mesh, so each mains-powered device extends the network instead of crowding it. Past roughly twenty devices this stops being theoretical.

You want battery devices that last. Wi-Fi is power-hungry. A Wi-Fi sensor eats batteries; a Zigbee or Thread sensor can run for a year or more on one. If you want door, motion, or leak sensors in any quantity, you want a low-power protocol, and that means a hub or a border router. Our Matter and Thread guide explains why the two aren't the same thing.

You want it to feel fast. Local control is noticeably quicker than a round trip to a server and back. The delay between pressing a switch and the light coming on is the difference between a home that feels solid and one that feels like a gadget.

You want to keep older kit. A hub can bridge existing Zigbee or Z-Wave devices into a modern ecosystem, which is usually cheaper than replacing them.

The trade-offs nobody mentions

A hub is a single point of failure. Everything routes through it. When it dies, or when its manufacturer discontinues it, you have a much bigger problem than a dead bulb. Favour hubs from companies likely to still exist in five years, and favour ones that keep working locally rather than depending on the vendor's servers just to function.

Setup is more work. Two systems to configure instead of one, and one more suspect when something misbehaves.

Cost. A dedicated hub is money not spent on devices that actually do something. For a small system that's a bad trade.

So, which way?

Under roughly ten devices, all Wi-Fi, no plans to expand. Skip it. Buy the devices. You can add a hub later, and by then you'll know whether you need one.

Growing past ten or twenty devices, or adding battery sensors. Get a hub, and get one with Zigbee and Thread radios rather than a Matter-controller-only device. The Echo Hub is the strongest option if you're already in the Alexa ecosystem, and it doubles as a wall-mounted control panel.

Outages genuinely matter to you, because of where you live or what you automate. Get a hub, and check specifically that it runs automations locally rather than merely calling itself a hub.

You already own a modern speaker or display. Check what it includes before spending anything at all.

Bottom line

Matter didn't kill the hub. It changed what a hub is for.

A hub used to be the thing that made your devices talk to each other, and that's the standard's job now. What's left is the part that was always the most valuable and the least advertised: a hub is what keeps your home working when the internet doesn't, keeps forty devices off your router, and keeps your sensors alive for a year on a coin cell. If none of that describes your house, you don't need one, and anyone insisting otherwise is selling a box.

Starting from nothing? Our beginner's guide covers picking an ecosystem before you buy any hardware, which is still the decision that matters most.