What your HR and payroll systems actually cost you — licences, admin hours, correction cycles and risk. About two minutes, and nothing is stored.
The bill you pay for HR software is the visible part. Errors, manual work, and risk are the rest — let's size the whole thing. Eleven questions, about two minutes.
Multi-country payroll is where penalty exposure multiplies — Alight found 67% of multi-country operations hit compliance penalties, against 24% of single-country ones.
Every state is its own employer registration — and withholding and unemployment are two separate accounts, not one. Each state then adds four to nine returns a year on top of the federal six.
Salary sets what an hour of everyone's time is worth, which is what makes self-service friction and turnover cost real money. Pay frequency sets how many chances a year there are to get it wrong.
All-in — subscriptions, support, add-ons, implementation amortised. A rough number is fine; it's the one line you can check against an invoice.
Count everything HR or payroll touches: HRIS, payroll, time & attendance, benefits, ATS, LMS, spreadsheets that people actually rely on. Every extra system is another reconciliation.
Be honest — this is the single setting that moves the number most. "About average" matches the typical company in the EY payroll-error study.
This is who absorbs the manual work. Answer in the numbers you actually have — salary and benefits — and we'll work out the fully loaded hourly cost ourselves. Leave anything blank and we'll use a benchmark.
Re-keying between systems, chasing approvals, correcting records, rebuilding reports in a spreadsheet. If you've never measured it, skip — we'll infer it from your rating on the last question.
Off-cycle runs, retros, manual adjustments, anything re-issued. This is the most valuable answer in the whole tool — give us a real count and we stop using the study average for your largest cost line and use your own error rate instead.
Fines, lawsuits and turnover are expected values — a probability times a cost, not a bill you've received. This dials only those three lines up or down. It never touches what you pay or what you can measure.
HR Systems Cost Assessment
People Street True Cost of HRIS™ Assessment · peoplestreet.co
Figures are rounded, so they will not always match the exact multiplication — this is an estimate and it says so. Every line's precise arithmetic is on the calculator page.
Your invoice — you enter it, or we prefill the published per-employee average until you do.
Measured waste: payroll and data errors getting found and fixed, scaled by how well you rated the system, plus your team's manual hours and re-keying between systems.
Risk math, not a bill: the chance of a fine, a pay dispute, or someone quitting over repeated pay errors, times what each costs when it lands. The only part the caution setting moves.
Cuts across two of the three. Every U.S. state beyond your first is a separate employer registration — withholding and unemployment are two accounts, not one — each with its own return cadence on top of the federal six. That adds staff hours and raises the number of agencies that can assess you. Where nobody is registering or filing for you, the exposure steps up again, because an unregistered employer can be looked back on further than a registered one.
Never asked for and never guessed. Wherever this model prices somebody's time it builds the rate from the salary you gave us plus statutory employer payroll taxes, your share of a health premium and your 401(k) contribution, then divides by paid hours. That lands below the ratio implied by the BLS employer-cost series, because we leave out state unemployment tax, paid leave and workers' comp and never net holidays or PTO out of the divisor. Deliberately conservative.
The error benchmark is published for larger U.S. employers and is applied here per employee, which is the conservative direction at your size: the fixed work of running payroll and compliance does not shrink in proportion to headcount, so a per-head rate understates it for a smaller team rather than flattering it. All figures USD.
People Street model assumptions, not measured findings. Change any — results update live and label themselves "Custom assumptions."
Ask your payroll team how many corrections went out last quarter — off-cycle runs, retros, voided checks. That single count moves this figure more than any other answer, and it takes one message to get.
Until you do, the software line is a published average rather than your bill, and the multiple against it stays blank. It is the one input that turns an estimate into your estimate.
Read the recommendation above. If most of what this found is recoverable where you are, a migration buys you the same money eighteen months later and costs you a year of disruption first.
How precise this is. It is a planning estimate, built from published research and the answers you gave — good enough to decide whether the problem is worth a proper look, not good enough to put in a board pack on its own. Turning it into a defensible number means going through your actual pay runs, your actual HR tickets and your actual contracts. Twenty-eight sources stand behind the model, held to two tiers: primary research verified against the publication carries the arithmetic, and practitioner benchmarks may only prefill a field you have not answered.
That is the work People Street does. If the figure above is roughly right, it is almost certainly worth an hour.
Let us check this against the real thing.
We take this estimate and check it against your real pay runs, your correction log and your HR tickets – so the figure stops being a planning number and becomes one you can defend.
Book 20 minutespeoplestreet.co/book
The figures stay on this page either way.
Your report is on its way. The branded version is unlocked on this page too — print or save it whenever you like.