Comparing payroll suppliers on one scenario, current HMRC status and full cost

A fair supplier comparison uses one employer scenario, the same evidence date and the exact product edition. A table mixing a free micro-employer tool, bureau platform and enterprise managed service will look comprehensive while answering no practical buying question.

Set a reference scenario

For example: one English limited company, 42 active employees, monthly payroll, variable overtime, directors, starters and leavers, one named pension provider, two approvers and an accounting-journal export. Replace those assumptions with the buyer's facts.

HMRC's find payroll software guide supplies feature prompts including payslips, pensions, different pay periods and reports. Add every ordinary and boundary case the employer actually uses.

Build the product table

Criterion	Evidence required	Do not accept
PAYE reporting	Exact product on current HMRC list	Supplier-wide badge
Tax-year scope	Versioned product and test record	Undated "compliant" claim
Pension workflow	Named scheme, task and accepted exchange	Generic pension tick
Migration	Reconciled year-to-date parallel run	Imported row count
Access	Role and audit test	Security adjective
Support	Channel, hours and exercised escalation	Award alone
Cost	Identical scenario and contract term	Promotional headline
Exit	Complete usable export and deletion process	Vague portability promise

HMRC's recognised product directory confirms online PAYE reporting for listed products at the time checked. HMRC says it cannot recommend suppliers. Capture product name and date rather than copying the whole page indefinitely.

Separate supplier statements from tests

For each candidate, record a first-party feature page, plan, access date and exact claim. Mark it "supplier stated" until verified. Use synthetic data to test calculation, permissions, submissions, pension files, corrections and export.

When a feature requires an add-on or higher plan, show that in the result and total cost. If a fact is available only through sales, request it in writing and mark it unverified until received.

Score suitability, not fame

Weight essential requirements before product research. A missing required pay frequency can eliminate a candidate; a larger template library should not compensate. Keep usability, implementation, support, security and commercial terms as separate dimensions.

The Pensions Regulator's software checklist is a useful independent prompt for pension tasks. Apply it to the actual employer and scheme.

Publish limitations

State that the comparison is a point-in-time assessment, not tax or purchasing advice. List products excluded, unavailable evidence, tests not performed and commercial relationships. Recheck recognition, tax-year support, integrations and price immediately before a decision.

The winning supplier is the one whose documented and tested product fits the scenario with acceptable residual risk and cost. That conclusion should change when the scenario or evidence changes.

Inspect implementation and service capacity

Ask who configures the product, validates migrated data and investigates differences. Record the proposed timetable, customer dependencies and number of parallel runs. A supplier should explain what happens if acceptance is not reached before the planned cutover.

Exercise support with a representative technical and payroll question. Note authentication, first response, escalation, resolution and evidence returned. Published opening hours are only one part of deadline readiness.

Review tax-year maintenance and end-of-life communication. Ask how the vendor identifies applicable HMRC changes, releases them, tests regressions and tells customers what action is required.

Record conflicts and commercial influence

Disclose affiliate fees, reseller relationships, free licences, sponsorship or consultancy that could affect the comparison. Keep paid placement separate from editorial scoring and apply the same source standard to every candidate.

Have a second reviewer reproduce decisive rows from the stored evidence. Where reviewers disagree, retain the underlying claim and resolve the interpretation before awarding a score. This process is slower than copying supplier tables, but it makes the conclusion auditable.