Payroll Health Score - methodology

Plain-English summary

The Payroll Health Score is a single number from 0 to 100 that summarises how clean and consistent your payslip looks. It is a roll-up - not a verdict - of around twenty categorical checks we run across tax, NI, statutory minimums, pension, and consistency.

Each check returns one of four states: pass, neutral, warn, or fail. We weight the checks and convert the weighted result into a 0-100 number. The same number then maps to one of four bands so you have a quick at-a-glance read.

For subscribers we keep the score over time. That trend is more useful than any single score: a 78 in isolation is fine; a slow drift from 88 to 78 over six months is the kind of pattern that is worth a conversation with payroll.

Categorical inputs

Step-by-step scoring logic

1. Run all checks. Each check produces a categorical state from the set above.
2. Map states to points. pass = 1.0, neutral = 0.7, warn = 0.4, fail = 0.0.
3. Apply weights. Multiply the point score by the family's weight.
4. Sum. Add weighted scores; max possible is 1.00.
5. Scale. Multiply by 100 and round to the nearest integer.
6. Apply caps. Any single fail in a 20%-weighted family caps the overall score at 74. Two fails cap at 49.
7. Band the result. Use the four bands below.
8. Persist (subscribers only). Store the score and per-family states for trend lines.

Banding

Worked example

Liam's April 2026 payslip produces:

• PAYE consistency: warn (0.4 × 20% = 0.080)
• NI consistency: pass (1.0 × 20% = 0.200)
• Tax-code plausibility: pass (1.0 × 10% = 0.100)
• Statutory minimum: pass (1.0 × 10% = 0.100)
• Pension auto-enrolment: neutral (0.7 × 10% = 0.070)
• Net reconciliation: pass (1.0 × 10% = 0.100)
• YTD continuity: pass (1.0 × 10% = 0.100)
• Statutory pay flags: pass (1.0 × 5% = 0.050)
• Slip completeness: pass (1.0 × 5% = 0.050)
• Sum: 0.850 → score 85 → Green band

Liam's previous three months were 88, 87, and 86, so the trend is mildly downward but within noise. The PAYE warn drove this month's drop and is highlighted as the priority question for payroll.

Trend history (subscribers)

For paid subscribers we retain the score and the per-family state for up to 12 months. We also store a rolling moving average and surface drifts of more than 5 points across a three-month window. Trend data is opt-in at account creation and can be exported or deleted at any time.

Edge cases handled

• Low-confidence OCR - checks that depend on the unreadable field are forced to neutral, not fail.
• Emergency tax codes - YTD continuity is forced to neutral.
• First slip with new employer - YTD continuity is forced to neutral.
• Bonus month - net reconciliation tolerance widens by 5%.
• Mid-year category change - NI consistency check uses the slip's stated category, not an inferred one.

Limitations and known gaps

• The score is a heuristic, not an audit. A 100 does not certify your payslip is correct.
• Weights are reviewed quarterly and may change. Historic scores are not retroactively recomputed.
• Caps mean a single material fail will always read amber or worse, even if the rest is clean - by design.
• The score does not factor in employer-side benefits or other off-slip considerations.

Sources used

Disclaimer

The Payroll Health Score is an indicative summary, not a regulatory or audit determination. PayslipIQ does not provide tax, legal, or employment advice. Use the score as a starting point for conversations with your employer or a qualified adviser.