The question behind the question
“Can you actually make a living off this?” The question keeps surfacing in the forums, sometimes flat out, more often in disguise. Someone writes that they’ve been running Home Assistant at home for years, that friends keep asking whether he couldn’t set it up for them too — and then the line that gives away the real hesitation: “But can you build a proper business on something like this? It’s open source, there’s nobody standing behind it.”
I find the question honest, and I get it. When you stake your income on something, you want to know what you’re leaning on. With a closed system that comes with a vendor, a hotline and a contract, that feels clearer at first glance. With Home Assistant there seems to be — well, nobody there.
Except that first glance is wrong. And the proof is over fifteen years old, runs on roughly forty percent of all websites on the internet, and is called WordPress.
Three fears, all built on the same pattern
When I talk to people toying with the idea of looking after HA installations professionally, almost the same three worries come up every time. Not as excuses — as real, legitimate concerns.
The first: there’s no vendor behind it. Who’s liable when something breaks? Who do I point to when the customer asks why his front-door lock has been dead since the update? With Loxone or Control4 there’s a vendor, a certification, an escalation path. With open source there’s a GitHub issue and, when push comes to shove, a shrug.
The second: I’ll have to be on site constantly. The idea that every fault means a drive out, because you can’t reach in from a distance, is ruinous for a business model. Spend forty minutes in the car per incident and you can’t look after twenty customers. Five at most, and even those only if they don’t all act up at once.
The third: the thing is unstable. Updates blow away configurations, a plugin breaks after a HACS update, a breaking change in Core kills an automation overnight. Plenty of people still file Home Assistant under tinkering platform — and you don’t build a trade on tinkering.
Three fears, one common denominator: the feeling of leaning on something that could give way beneath you at any moment, with nobody there to catch you. That exact feeling has existed before. In another ecosystem, with a strikingly similar history.
WordPress had precisely this reputation
Rewind to, say, 2010. WordPress was open source, free, carried by a community. And it had exactly the reputation Home Assistant has today. “Updates wreck the site.” “Plugins are a security nightmare.” “No official support, you’re on your own.” “For a serious business you use something real, not some hobby thing off the internet.”
Suggest to a mid-sized company back then that they build their corporate site on WordPress, and you had to justify yourself. Today it’s the other way round — the one who doesn’t suggest it has to explain why.
What happened? Not that WordPress turned stable overnight and some corporation took on the liability. Something else happened: a whole industry grew up around people who run WordPress professionally. Agencies, freelancers, one-person shops who don’t build a site and wave goodbye but tend it for the long haul. Apply updates, run backups, respond when something breaks — for money, monthly, predictable. The famous maintenance and care plans that today form the backbone of countless agencies. Recurring revenue from software nobody pays a licence fee for.
This is the point where I like to pin down aspiring HA integrators. The missing vendor isn’t a hole. It’s the basis of the business. Precisely because no hotline exists, the customer needs someone to fill that role — and that someone is you. “No vendor behind it” reads, from the customer’s side, as “I need a person I can trust.” That isn’t a deficit. That’s demand.
The part most people miss: the tooling
But — and this matters to me, because the response here is too often a breezy “well, just do it then” — the WordPress industry didn’t grow on goodwill alone. It became possible because at some point the tools arrived to tame the madness.
Picture a service provider with eighty WordPress customer sites. Eighty logins. Eighty plugin lists. Eighty backup routines, eighty update states. Anyone logging into each site individually every morning to check whether the night broke something gives up after two weeks. It doesn’t scale. It’s the on-site fear of the WordPress world, just in digital form.
The thing that solved it is a class of software people call management dashboards. The best-known free example is MainWP — self-hosted, open source, built for precisely this pain. You install a small extension on each customer site, and from then on you see everything from a single dashboard: which site needs an update, where a plugin is out of date, which one is currently offline, when the last backup ran. Updates for core, plugins and themes across every site at once. Uptime monitoring. Security checks. Reports you send the customer so they can see what they’re paying for.
Instead of logging into eighty sites, the provider glances at one board and sees where action is needed. Whatever needs remote access, he handles remotely. The drive out — the great bogeyman — shrinks to the handful of cases where someone genuinely has to put hands on the hardware.
This is exactly the tool that was missing for Home Assistant for a long time. And as long as it’s missing, the on-site fear stays justified.
HA Fleet Manager is the MainWP moment for Home Assistant
Here’s where the product behind this blog enters — and I’ll spell out deliberately what it is and what it isn’t, because marketing speak around this topic gets on my own nerves.
HA Fleet Manager is meant to be for Home Assistant what MainWP is for WordPress. A central dashboard for many instances. You install a custom integration on each customer installation, and after that you see the state of the whole portfolio at a glance: which instance is online, which integrations are running, which HACS plugins are installed at which version, whether critical errors are sitting in the logs. Remote access to the HA frontend when you need to step in — but only after the customer has granted it, and only for a time-boxed window. A maintenance history that doesn’t live in your head.
The parallel holds right down to the detail. The on-site fear dissolves in the HA world for the same reason it did back in the WordPress world: because you no longer have to drive out to every installation or log in one by one to know how it’s doing. You see it. And when something’s up, you step in remotely. What that looks like day to day I’ve played through elsewhere as a small story — the Z2M cascade on a Thursday afternoon, three customers, forty minutes, no drive out.
What HA Fleet Manager is not: a finished, decades-hardened product with the feature surface of an established RMM suite. It’s young. It’s being built right now. But MainWP was young in 2013 too, and WordPress more so in 2010. Ecosystems mature once somebody starts building the tools the pros need.
That leaves the third fear: stability
We’ve covered the missing vendor and the on-site worry. The third — “the system fails because updates break things” — deserves its own honest answer. Because unlike the other two, it can’t be waved away with “WordPress managed it as well.”
First: yes, updates sometimes break things. That’s true and stays true. But that’s precisely the professional lever, not a disqualifier. The difference between a hobbyist and a service provider isn’t that nothing ever breaks for the provider. It’s that the provider notices first — before the customer calls. Roll an update out in a controlled way on a pilot installation, watch it, and only then push it to the fleet, and you’ve turned “something broke” into a planned operation. That’s exactly what a fleet dashboard is for: not to prevent faults, but to see them early and all in one place.
Second: the “tinkery” reputation is lagging behind reality. Home Assistant left the basement corner long ago. There’s a Foundation, a fixed release cadence, the “Works with Home Assistant” programme, official hardware. This isn’t a weekend script any more, it’s a platform with millions of installations. Anyone still saying “hobby thing” today is describing the state of five years ago.
And third, the point closest to my heart: with open source, stability isn’t a property the vendor ships. It’s a service the provider delivers. That sounds like more responsibility — and it is. But it’s exactly the responsibility the customer is paying for. A maintenance contract doesn’t sell “the software never breaks”. It sells “when something’s up, somebody takes care of it, and usually before you even notice”. A vendor with a closed system is often worse at delivering that than a human being who knows those twenty installations inside out.
What this means for you when you start
I don’t want to talk anything up here. Building an HA business is work, and it’s no guaranteed runaway success. But the excuse “can’t be done, it’s only open source” has been disproven by history. It was disproven once already in a structurally very similar market, thoroughly, by people who make a good living off it today.
What it takes is the same as back then: the willingness to fill the role of the missing vendor yourself — as a trusted person, not a tinkerer. A business model with recurring revenue, so you’re not living from emergency to emergency. And a tool that doesn’t force you to log into twenty installations one by one every morning. What this profession actually looks like, from the typical day to the hourly rates, I’ve described in more detail in the HA integrator job profile. And if you’re still undecided which tool fits your scale, the comparison of six ways to manage multiple HA instances might help.
The WordPress crowd in 2010 had no MainWP and no certainty their industry would even come to exist. They just started. Today you start, and the tool already exists.
That, honestly, is a pretty good moment to begin.
Disclosure: HA Fleet Manager is the product behind this blog. The WordPress parallel isn’t a sales gimmick all the same — it’s the reason I believe this market works.