Before you begin

Every answer is a fact: a count, a percentage, or a choice describing what actually happens. The model computes your levels from the facts — you never rate yourself. Answer for the ordinary hard month, not the best one. Basics below are context and denominators — never scored.

Your answers save as you go on this device. You can come back to the same link. Your written assessment arrives by email shortly after completion.

Organisation *

Industry *

Headquarters country *

Employees on payroll *

Countries you run payroll in *

Pay frequency * Weekly Bi-weekly Semi-monthly Monthly Mixed

Payroll runs per months (incl.off-cycles) *

Operating model * In-house Co-sourced Outsourced Hybrid PEO / umbrella

Payroll system(s) in use *

Your answers are used to generate your report and, in anonymised and aggregated form, to build the Modern Payroll Maturity Model benchmark. We will never publish anything that identifies your organisation.

Does a documented readiness sign-off happen before processing starts? *

There is no readiness check An informal check - depends who is running A documented checklist - most runs A documented sign-off - every run A documented sign-off every run - processing cannot start without it

In a normal cycle, how many trial runs happen before the live run, and does a reviewed variance register gate it? *

No trial run One trial - register not always reviewed One trial - register reviewed and signed Two trials with reviewed registers Two or more - signed register gates the live run

How many of these reconciliations run every cycle: headcount-in vs payslips-out; gross vs prior run; register vs bank file; register vs GL? * Number 0-4

How many separate people must review and sign payroll results before payment is released, and how formal is it? *

One person Two - informally Two - formally defined Three including independent QA Three or more including statutory review - system-enforced

During processing, how are unusual results caught? *

By eye - if at all Spot checks on known trouble spots Automatic checks on some pay elements Automatic checks on most pay elements Automatic checks across the run - flagged with reasons and owners

Can someone who edits payroll data also release the bank file? *

Yes Yes - but it would be noticed later No - by policy No - blocked by the system No - system-blocked plus dual bank approval

Within how many working days of pay day is the register-to-bank-to-GL reconciliation completed and signed? *

Not done Not every month By month-end Within 1-3 days Same day

When was payroll system access last reviewed, and what did the review cover? *

Never or cannot say Over a year ago Within the year - user list only Within the quarter - users and roles Quarterly - roles and segregation conflicts with removals evidenced

How does pay data move to third parties (bank, bureau, auditors) today? *

Unencrypted email attachments Password-protected files by email Mostly secure channels - some email Secure channels only Secure channels only - with access logging and retention rules

How do payroll inputs mostly arrive? *

Almost everything arrives by email or spreadsheet Mostly files and spreadsheets - some interfaces About half through system interfaces Mostly through interfaces - the rest on standard templates Nearly everything system-to-system - validated on entry

How often do master-data problems (bank details, IDs, eligibility) delay a run or force corrections? *

Almost every run Most months A few times a year Rarely - once a year or less Practically never - master data is validated before it reaches payroll

When people join or leave, does their first or last pay land right without chasing? *

Usually needs chasing and fixing Often needs chasing About half go straight through Mostly right without chasing Almost always right - exceptions are rare and logged

How many of these input types are validated automatically at entry: headcount changes; compensation changes; bank details; one-off payments; attendance? * Number 0-5

When payroll catches an input error, what usually happens? *

Payroll patches it and moves on Payroll patches it - the source hears about the big ones About half get fixed back at the source Most get fixed at the source (HR, time, finance) Source fixes are the rule - repeat errors are tracked to zero

How many of your core systems (HR, time and attendance, finance or GL, banking) exchange payroll data automatically with the payroll system? * Number 0-4

Does the payroll cut-off hold? *

There is no real cut-off - inputs arrive until pay day A cut-off exists but usually slips Holds most months Holds except for rare approved exceptions Holds every run - late items wait for a governed off-cycle

Which payroll run types (regular, off-cycle, exit, bonus) run on your main platform rather than outside it? *

Most run outside the platform Regular runs only Regular plus some other types All types except rare exceptions All types - under the same controls

How many off-cycle or correction runs did you need last quarter, beyond the planned calendar? * Number of runs

Does everyone get paid on the promised day? *

Pay day slips or misses happen most months Some employees miss pay day a few times a year Occasional individual misses - promptly fixed No one has missed pay day in the last year No one misses pay day - and we can evidence it run by run

How are payslips and payroll outputs delivered to employees? *

Printed or emailed manually Emailed from the payroll mailbox Self-service portal for most - manual for some Self-service for all - secured Self-service for all - secured with delivery tracked and access logged

How heavy is the payroll query load after pay day? *

The days after pay day are a helpdesk A steady stream all month A manageable bump after pay day Light - queries are tracked and falling Minimal - self-service answers most and the rest are tracked to SLA

Is payroll accuracy measured with a defined denominator, and what was last quarter's error rate? *

Not measured Estimated informally Measured - worse than 1 error per 100 payslips Measured - between 1 per 100 and 1 per 1000 Measured - better than 1 per 1000

Is on-time payment measured against a defined SLA, and what was it last quarter? *

Not tracked Tracked informally Tracked - below 98% on time Tracked - 98 to 99.5% Tracked - above 99.5%

Is completeness checked each run (everyone due pay is paid; every input is processed)? *

No Only when something goes wrong Headcount check each run Headcount and input completeness each run Both - reported against target monthly

What was your first-time-right rate last quarter (runs approved with no rework after first submission)? *

Not measured Estimated informally Measured - below 75% Measured - 75 to 90% Measured - above 90%

What was the median time to resolve an employee payroll query last quarter? *

Not tracked Tracked informally Over 5 working days 2 to 5 working days Under 2 working days - tracked against an SLA

How many of these exist as regular reports: cost run-over-run; error and rework trend; off-cycle trend; input quality by source; query volumes? * Number 0-5

Between inputs arriving and the engine running, how much manual handling happens? *

Everything is re-keyed or massaged in spreadsheets Heavy - many touchpoints every run A handful of known touchpoints One or two - documented None - inputs flow straight through and exceptions are logged

How often does completing payroll take evenings or weekends? *

Every run Most runs A few times a year Rarely - year-end at most Never - the run fits the working day

When something needs correcting mid-run, what happens? *

Someone edits the data directly Direct edits - usually mentioned to someone Logged informally and approved after the fact Logged and approved before the change - mostly Every correction: logged, approved, rerun - no exceptions

When an error is found, what happens after the fix? *

We fix the payslip and move on The big ones get discussed Root cause recorded for major errors Root cause recorded for most - fixes tracked Every error: root cause, preventive fix, trend reviewed monthly

If your payroll lead resigned today, how much of your core process documentation is current (updated within 12 months)? *

No usable documentation A few documents - mostly stale About half current Most current - reviewed annually All current - versioned and reviewed on schedule

How often does a named owner review payroll performance with the team (and the provider, if any)? *

No such review Only when something breaks Quarterly Monthly Monthly - with actions tracked to closure

Is there a written escalation path for payroll incidents, and when was it last used or tested? *

No Informal - people know who to call Written - never tested Written - used within the last year Written - tested with post-incident reviews

How is payroll performance by your provider or internal team evidenced each month? *

No regular reporting Verbal updates only Reports received but not questioned Reports reviewed against SLA with actions agreed Evidenced pack reviewed - actions tracked and consequences applied

If pay failed for part of your workforce tomorrow, is there one named owner for the incident end to end? *

No clear owner Depends on where it fails Named owner by policy - never tested Named owner - tested in the last year Named owner with cross-functional authority - rehearsed and reviewed

Of the statutory payments and filings due in the last 12 months, how many were on time and evidenced? *

Cannot say Some were late All on time - evidence scattered All on time - evidence filed All on time - evidenced with sign-off before submission

In how many of the countries you run payroll does a named person, internal or provider-side, own statutory compliance end to end? *

No named owners Some countries only All countries - via provider with no verification All countries - verified annually All countries - verified with evidence every cycle

After each run, is statutory withholding independently validated before filing? *

No Spot checks only Every run - manually Every run - tool-assisted Every run - tool-assisted with exceptions logged and cleared

Of the regulatory changes affecting payroll in the last 12 months, how many were implemented and verified before their effective date? *

Cannot say Some missed their date All landed - some verified late All landed on time All on time - via a tracked pipeline with one owner

In a normal run, how many employees process straight through with no manual touch? *

Hard to say - most need some handling Fewer than half Most - maybe three in four Nearly all - only defined exceptions are touched All but a logged exception list - and we know the exact rate

Where do your pay calculations actually live? *

Key calculations live in spreadsheets Several side calculations every run A few known side calculations One or two edge cases outside the engine Everything computes in the engine - no side calculations

Where is AI in your payroll operation today? *

Nowhere - and the foundations are weak Exploring - foundations still weak Foundations building - no AI yet Piloting - human-reviewed and auditable In production - governed, explainable, auditable

If an automated or AI-driven step mis-paid someone tomorrow, could you detect it, explain it, and reverse it? *

Could not detect it It would surface when employees complain Detected after pay day - explained with effort Detected before pay day - explainable Detected - explained and reversed, rehearsed and logged

How much of gross-to-net (proration, arrears, statutory) does the engine compute by rule? *

Much of it is worked out manually The engine does the basics - people do the rest The engine does most - regular manual workings remain The engine does nearly all - rare manual workings All of it - by rule, every run

How does the team review a run before approving it? *

The whole register - line by line The whole register - sampled Flagged exceptions plus a register scan Flagged exceptions only - most runs Exceptions only every run - the flags are trusted and tested

How does payroll improve? *

It doesn't - there is no time Fixes happen after incidents A few improvements a year - informally A steady pipeline - some measured A managed pipeline - each improvement measured for effect

Is there a defined payroll close each period, ending in a signed pack (reconciliations, exceptions, MIS)? *

No defined close Close happens - nothing signed Signed close most months Signed close every month Signed close every month - pack archived audit-ready

How did your last year-end close actually go? *

Major rework or restatements Late - heavy manual effort On time - heavy manual effort On time - mostly routine A non-event - reconciled all year and closed on schedule

If an auditor picked one payslip from last quarter, could you show every change, approval and calculation behind it? *

No Partially - with real effort Yes - with a day's notice Yes - same day Yes - on demand, system-logged end to end