Ask who owns your HR system and you will get a confident answer from three different directions, each of them wrong in the same way. HR will tell you it is an IT system — they configure it, they run the servers, they hold the vendor contract. IT will tell you it is an HR system — it is HR's processes, HR's data, HR's people who live in it every day. Finance signed the cheque and considers ownership someone else's problem entirely. Everyone touches it. Everyone depends on it. And in the specific sense that matters, no one owns it.
This is not a personality problem or a turf war that better manners would solve. It is structural. An HR system sits precisely on the seam between three functions that each own part of it and none of it — and a thing owned in parts by several groups is, in practice, owned by nobody. The system does not fail loudly when this happens. It just quietly stops getting better, because improving it is nobody's actual job.
The implementation had an owner — a sponsor, a project manager, a team with a mandate. The moment it goes live, that ownership evaporates, and unless someone deliberately replaces it, the system enters a long ownerless drift that most organizations mistake for stability. It is not stability. It is neglect that has not yet produced a visible failure.
Owning the Technology Is Not Owning the Outcomes
The reason "IT owns it" feels like a clean answer — and isn't — is that it quietly swaps one kind of ownership for another. There are two, and confusing them is the root of the whole problem.
The first is ownership of the technology: keeping it running, patched, integrated, and available. Uptime, access, security, the vendor's technical relationship. IT can own this cleanly and often does it well. It is real work and it matters.
The second is ownership of the outcomes: whether the system actually serves the business. Whether the processes it runs still match how the organization works. Whether the data is trustworthy. Whether the thing is getting better or slowly rotting. This ownership has no natural home on the org chart, because it is neither purely technical nor purely HR — it is the judgment that sits between them. And it is precisely the ownership that goes missing, because everyone assumes "IT owns it" covered it. It didn't. IT owning uptime tells you nothing about whether anyone owns whether the system is any good.
What an Ownerless System Looks Like
The absence of an owner is invisible for a while, because nothing breaks. Then the symptoms arrive, and they are always the same, because they are all the same problem wearing different clothes.
The change request that needs HR to define it, IT to build it, and Finance to fund it — and so sits for six months, because it needs three groups who each believe it is another group's call to move first. The data quality that degrades because no one is accountable for it, only for their slice of it. The vendor relationship that goes unmanaged between renewals, so you discover your roadmap only when the invoice arrives. The module you paid for and never turned on, because switching it on was no one's job. The workaround that becomes permanent because fixing it properly requires an owner to decide it is worth fixing.
None of these is a crisis on its own. Collectively they are the signature of a system that has no one steering it — an asset the organization is maintaining but not improving, paying for but not developing. It works. It simply never gets better, and slowly the gap between what the business needs and what the system does widens until someone proposes replacing it — which is how a perfectly good system gets thrown away for want of an owner.
Ownership Is a Role, Not a Committee
The instinct, once an organization notices the gap, is to fill it with a committee. A governance forum, a steering group, a cross-functional council where HR, IT and Finance meet monthly to jointly own the system. This feels responsible and changes nothing, because it re-creates the original problem in a conference room: a thing owned by everyone in the meeting is still owned by no one after it.
Committees are useful for deciding. They are useless as owners, because ownership is the thing that happens between meetings — the single point of accountability who holds the roadmap, arbitrates the trade-offs, manages the vendor, and answers for whether the system is serving the business or drifting from it. A forum can advise that person and hold them to account. It cannot be that person. Shared accountability is a polite phrase for none, and every system run by a committee eventually proves it.
The Role This Keeps Pointing Back To
Issue 02 of this newsletter described the role nobody budgets for — the HR Tech expert who bridges business operations, HR processes and technology decisions. Ownership is the permanent, post-go-live form of that role. The bridge that was useful during selection and vital during implementation becomes, once the system is live, the answer to the ownership question itself — because the owner has to live in exactly the space between HR and IT that the bridge role occupies.
That person owns the outcomes, not just the technology. They hold the roadmap and decide what the system does next. They speak both languages well enough to translate a business need into a technical change and a technical constraint into a business trade-off. They manage the vendor as a relationship rather than a renewal. And they are one accountable person, supported by a governance forum but not replaced by it. Without that role, an HR system does not fail — it simply drifts, expensively, until someone mistakes the drift for a reason to start the whole cycle over again.
A Final Thought
The question "who owns the HR system?" sounds administrative, the kind of thing that gets settled in an org-chart footnote. It is actually the question that determines whether the investment you spent a year making keeps returning value or slowly stops. A system with an owner gets better; a system without one gets replaced. The difference between those two outcomes is not budget or technology. It is whether one identifiable person wakes up accountable for whether the thing is any good.
Getting it right does not require a reorganization. It requires naming an owner — a real person, not a committee — who owns the outcomes rather than the uptime, who holds the roadmap and the vendor and the trade-offs, and who has the mandate to actually decide. Do it deliberately at go-live, before the ownerless drift sets in, because the cost of not doing it is invisible right up until the day someone proposes spending another year and another budget replacing a system that only ever needed someone to own it.
A system that belongs to everyone is a system that belongs to no one. Give it an owner, or watch it slowly become nobody's problem until it becomes everybody's.
Owning the uptime is not owning the outcomes. Someone can keep the HR system running perfectly while no one at all is accountable for whether it's any good — and that gap is where the value quietly leaks out.