Every HR system starts with a business case. It is a serious document — savings quantified, efficiencies projected, better decisions promised, all of it totted up into a number large enough to justify a very large cheque. It is argued in front of leadership, defended in a budget meeting, and, on the strength of it, the money is approved and the project begins. It is often the most rigorous piece of financial thinking the whole initiative ever receives.
And then, in almost every organization, it is never opened again.
The business case did its one job — it unlocked the budget — and the moment it did, it was quietly retired. Nobody goes back a year after go-live to ask whether the savings materialized, whether the efficiencies appeared, whether the decisions actually got better. The document that justified the entire investment is filed away the instant it succeeds, and the promise it made is never once checked against what happened.
The result is that organizations spend enormous sums on HR systems, run them for years, and genuinely cannot tell you whether the investment paid off. Not because it failed — it may well have succeeded handsomely — but because no one ever went back to measure. The single most expensive decision the function makes is also the one whose outcome it most reliably declines to examine.
The Business Case Was Written to Win, Not to Keep
The reason it never gets revisited is baked into what it is. A business case is not, despite appearances, a measurement instrument. It is a persuasion instrument — a sales document, pointed inward at your own leadership, whose purpose is to get to yes. That purpose shapes everything about it. The benefits are estimated optimistically, because a conservative case does not win approval. The costs are scoped to what is defensible, not to what will actually be spent over the life of the system. The numbers are chosen to clear the hurdle, not to be lived up to afterward.
None of this is dishonest; it is simply what any document designed to win a decision looks like. But a document built to win an argument is not built to be held to account, and everyone involved understands this at some level — which is precisely why no one expects to be asked about it later. The case was always going to be graded pass or fail once, at the budget meeting, and never again.
A number written to win approval and a number written to be measured against are two different things. The tragedy is that organizations write only the first kind, and then, having no second kind, conclude that measuring value is simply too hard to do.
"The System Works" Is Not "The Investment Paid Off"
When you do ask an organization whether its HR system was worth it, the answer you usually get is some version of "yes, it's working well." It runs reliably, people have adopted it, the data is decent, the vendor is fine. All of that is real, and all of it answers a different question than the one that was asked. Operational success and financial return are not the same thing, and conflating them is how the value question gets quietly dodged.
A system can run flawlessly — perfect uptime, high adoption, clean data, a smooth vendor relationship — and still have delivered little of the value that justified buying it. The business case did not promise that the system would work; working was assumed. It promised specific outcomes: this many hours saved, this cost avoided, this decision improved, this risk reduced. A system that runs beautifully but delivers none of those has succeeded operationally and failed financially, and only one of those two verdicts is the one leadership actually approved.
This is the substitution that lets everyone off the hook. "Is it working?" is easy to answer and almost always yes. "Did it deliver the value we bought it for?" is hard to answer and often uncomfortable — so the first question gets asked in place of the second, and the comfortable answer is quietly accepted as though it settled anything.
You Cannot Measure Improvement You Never Baselined
There is a practical reason the value question feels impossible to answer, and it is worth naming plainly: most organizations destroyed their ability to answer it before go-live, by never recording where they started. The business case claimed the system would cut the time to process a new hire, or reduce payroll errors, or speed up approvals. To prove any of that after the fact, you need to know what those numbers were before — the baseline. And the baseline is almost never captured, because at the point it could have been, everyone was busy building the system, not measuring the old one.
So a year later, when someone finally asks whether onboarding got faster, there is no honest answer available: not because it didn't improve, but because no one wrote down how slow it was to begin with. The measurement had to happen before the thing you were measuring changed, and that window closed at go-live, unnoticed.
This is why value measurement cannot be bolted on afterward as a report. It has to be designed in from the start — the handful of metrics the case actually rests on, captured as a baseline before the system changes anything, and then measured again on a cadence after. It is not a large undertaking. It is simply one that has to be started early, and someone has to own starting it, because by the time its absence is felt the opportunity to fix it is already a year gone.
A Final Thought
The uncomfortable truth is that most HR functions cannot prove the value of their largest investment, and that this is a choice — an accumulation of small, reasonable-seeming decisions to not capture the baseline, to not revisit the case, to accept "it's working" as the answer. It is a choice with a cost, and the cost lands at the worst possible moment: the next time the function goes to leadership to ask for money, and is asked, reasonably, what became of the last large sum it was given. "The system works" is a weak answer to that question. "Here is the value it delivered against what we promised" is the answer that funds the next thing.
Getting it right does not require turning HR into a finance function. It requires treating the business case as a living commitment rather than a spent argument: choose the few metrics the investment actually rests on, capture their baseline before go-live while you still can, measure them on a cadence afterward, and be willing to report the honest result — including where the value fell short. An organization that does this can say what its systems are worth. One that does not is left asserting that its most expensive decisions probably worked out, which is not the same thing, and everyone in the room knows it.
A business case that is never revisited is not a measure of value. It is a receipt for a promise no one intends to check.
A system can run flawlessly and still deliver almost none of the value that justified buying it. "Is it working?" and "Did it pay off?" are different questions — and only the first one is comfortable to ask.