Revenue-share audit: reproduce the monthly statement

Define the question and the authority

Decide whether you are checking one invoice, a configuration change or a recurring control. Those require different scopes. Record the agreement, amendments and effective dates before reviewing data. If a commercial email changes a rate, establish that it is an authorized amendment rather than assuming every message modifies the operative terms.

Our editorial audit method is a management review, not a statutory audit or assurance opinion. Assign an independent reviewer where practical: the person who configured a deal should not be the only person approving its calculation. Set an investigation threshold appropriate to the business, but still log systematic small errors that could accumulate across many partners or periods.

Trace eligibility and the NGR bridge

Rebuild the population using the permitted attribution and eligibility records. Confirm that the correct brands, markets and products are included. Next reconcile the aggregate revenue waterfall, checking each deduction against the agreement. SISU's terms illustrate explicit fee definitions and product aggregation; Pala's terms also contain program-specific NGR language. Their differences are a reason to avoid universal formulas.

Choose sample records that challenge the calculation: a standard record, a negative result, a reversal and a period-boundary event. Trace them from source to statement. The sample tests the path, while the aggregate bridge tests total completeness. Neither alone provides the same evidence as both, and a passed sample should not be described as proof that every record is correct.

References: SISU Partners — Terms and Conditions; Pala Partners — Terms of Participation

Reproduce a synthetic statement

Synthetic example: eligible NGR is €18,000 and the agreement carries €3,000 of negative NGR into the same pool. The remaining commissionable base is €15,000. A synthetic 30% rate produces €4,500. An approved €200 prior-period commission correction brings the statement to €4,700 before any separate settlement conditions.

The €200 correction is expressed in commission units, so it is added after the percentage calculation in this invented model. Adding it to NGR instead would produce only €60 of additional commission. This is why every adjustment needs a unit and a place in the formula. The example demonstrates an audit test and does not describe any cited program's actual calculation.

Test rate tiers and effective dates

A tier may apply one rate to the entire qualifying base or different rates to successive bands. Those approaches produce different results, so ask for an explicit worked example. Also establish which activity determines the tier, the period used and whether reversals can change it retrospectively. Do not rely on a dashboard badge that shows only the final percentage.

Test the value immediately below, exactly at and immediately above each boundary. Include a rate amendment halfway through the period if your arrangement allows one. The expected treatment should be written before running the test. If the platform cannot reproduce it, classify the issue as a configuration limitation or calculation defect rather than silently changing the expected commercial outcome.

Inspect adjustments and settlement separately

List every manual adjustment with its original period, reason, evidence, author and approval. A generic “finance correction” is not enough to reproduce the calculation. Review whether closed periods can be reopened and how affected partners are informed. Preserve the original and revised statements so the total change can be reconciled without relying on a person's memory.

Then inspect invoice and payment status. Minimum payout thresholds, incomplete supplier onboarding or disputed items can affect receipt without changing earned commission. Keep those causes outside the revenue waterfall. Match payment references and currencies to the approved statement, and have treasury explain any bank or conversion difference. This prevents settlement issues from being misdiagnosed as an NGR error.

Turn findings into durable controls

Write each finding as a condition, evidence, financial effect and corrective action. Assign an owner and retest date. A corrected spreadsheet is not a complete fix if the live platform will produce the same error next month. Check that configuration, contract documentation and reporting definitions now agree, then retain the retest evidence.

For renewals, summarize unresolved limitations in reporting access or reproducibility. Lack of evidence does not prove underpayment, but it limits confidence in the calculation. Use that distinction in the commercial conversation. The downloadable worksheet creates an evidence index for future reviews, with restricted access to any customer-level data and a separate approval trail for changes that affect payable amounts.

Working template

Download the CSV worksheet

Frequently asked questions

Sources and scope

Sources checked September 22, 2026. Numerical examples are illustrative unless explicitly identified otherwise.

• SISU Partners — Terms and Conditions — An example of defined deductions and aggregation across products. Checked September 22, 2026.
• Pala Partners — Terms of Participation — Program-specific commission, qualification and cap language; written account terms need checking. Checked September 22, 2026.

Related reading

• Build the deduction waterfall
• Resolve invoice and settlement variances
• CPA vs revenue share vs hybrid: compare the same cohort
• What counts as a qualified FTD in an affiliate contract?