How to Build the Internal Case for Payroll Operations Budget
Payroll loses the budget argument most years, and rarely because the case is weak. How to build one that survives a room full of functions with dashboards.
Every autumn, in most firms, a spreadsheet is opened that decides what each department gets next year. Payroll usually does badly on it.
Not because anybody is against payroll. Because that spreadsheet rewards whoever can show their working, and payroll is the one function whose entire output is the absence of a problem.
This article is for the person who has to make that ask. It assumes the case is sound and the problem is the evidence.
Why the payroll ask loses
Three reasons, and only one of them is about money.
The ask is made in adjectives. Busy. Stretched. At capacity. Every one of those is true and none of them is a number. The functions competing for the same budget arrive with charts, and a chart beats a strong feeling in that room every time.
The request is for headcount when the problem is drag. Asking for another payroller invites the question of whether the current team is being used well, which is difficult to answer without data. Asking to remove a specific, quantified inefficiency does not invite that question at all.
The risk of saying no is invisible. Every other function can describe what happens if it is refused. Payroll’s answer is that it will absorb it, because payroll always absorbs it, and everybody in the room knows that. A request that will be absorbed if refused is the easiest one to trim.
What to count, and for how long
A useful case needs one full cycle of evidence. Two is better. Four things:
Change volume, by client
How many payroll changes arrived, from whom, and how many were incomplete on arrival. Incompleteness is the number that does the work, because it converts a vague complaint about client behaviour into a measurable cost.
Chases
How many times the team asked for something that had already been promised. This is the purest measure of avoidable work in a payroll operation, and it is almost always higher than anyone expects.
Queries
Client questions answered, by client. In most operations this workload is completely unrecorded, which means the team is staffed for the payrolls that can be seen and quietly absorbs the helpdesk that cannot.
Time from arrival to resolution
Not to monitor anybody. Because it is the clearest available proxy for friction, and because it is the number that identifies which clients are heavy rather than simply large.
How to frame the ask
Three sentences, in this order. It works because it moves the conversation from whether payroll deserves more to what the firm is currently paying for the absence of a decision.
One: here is what the operation handled. A count, not an impression. Eleven thousand changes, four hundred of them after cut-off, across ninety clients.
Two: here is what it cost to handle it that way. The chases, the re-keying, the queries, expressed in hours. A day a week across a team of eight is a sentence with consequences.
Three: here is what that time is worth, and what releasing it would allow. Capacity for a defined number of additional payrolls, without a hire. This is the sentence that turns a cost request into a growth one, and it is the only version of the ask that is interesting to a partner.
What to leave out
Do not lead with the team’s stress. It is real, it matters, and it will be heard as a wellbeing issue rather than a commercial one. Lead with the commercial case and let the wellbeing benefit follow as a consequence.
Do not ask for a platform. Ask for a specific outcome and name the platform as the means. Nobody approves software. People approve capacity, risk reduction and margin.
Do not present a number you cannot defend. One count you are certain of beats five estimates. The moment a single figure is challenged successfully, the rest of the case goes with it.
The evidence problem, and the order to solve it in
There is a circularity here that stops most payroll teams before they start. The case needs data. The data does not exist, because the work happens in inboxes, phone calls and memory. And the system that would generate the data is the thing being asked for.
The way out is to gather one cycle manually. It is tedious and it is finite. A tally of chases and queries kept by hand for four weeks is enough to build a first case, and it is considerably more evidence than payroll usually brings.
After that, the record needs to build itself, because any manual tally is abandoned in the first busy week and rightly so. That is the practical argument for an operational layer around payroll rather than a spreadsheet beside it: every change captured where it arrives, every chase logged, every approval time stamped, as a by-product of the work rather than a task on top of it.
“With Changepen at our core, we’ve been able to take on 200 new payrolls without adding staff.” Martyn Cheney, Cheney Payroll Services. That is the shape of the sentence a business case is trying to earn: capacity gained without cost added.
If you want a quick view of where your own operation currently leaves a record and where it does not, our two-minute payroll service quiz is a reasonable place to start. No contact details required.