Two reports disagree about which affiliate deserves credit for a sale. That disagreement is worth investigating, but it does not yet tell you that anyone was overpaid. Start by finding out what each report counts and how it assigns credit. Otherwise, a reporting difference can become a commission decision before anyone has explained it.
Know which question the report answers
Google Analytics describes attribution as assigning credit to touchpoints along a path to an important action. Separately, Google Ads describes Conversion Lift as a controlled experiment comparing groups exposed and not exposed to ads. That is a concrete illustration of different measurement questions. Google's product documentation does not establish that the same experiment is available for a particular affiliate account; Google also says Conversion Lift is not available to every Ads account.
Follow one order through the credit rules
Imagine a fictional shopper reading a product review, later using a cashback link and placing one $100 order. Under an illustrative last-touch rule, the cashback partner receives all $100 of attributed order value. Under an invented equal-split rule for these two touches, the report assigns $50 to each partner. There is still one $100 order. These are fictional credit rules for explaining the arithmetic, not a claim about available Google Analytics models or a real customer journey.
The split report makes the review touch visible. It does not show what the shopper would have done without either partner. Nor does it determine which commission the actual agreement requires. Keep three records distinct: the order, the reporting credit and the payable commission. Adding $100 from one model to $100 from another would count the same order twice.
Reconcile the records before judging the partner
Use the reporting information your team is authorized to review. Compare the period, time zone, order status, adjustments and credit window. Check whether one report includes a later return or uses a different date for the same event. Ask the provider to explain any unresolved field or rule. Do not assume every network supplies a complete raw path export, or that an absent touch proves a publisher had no influence.
The earlier article attributed broad findings to media investigations and claimed incremental acquisition costs were commonly 20–40% higher than reported CPAs. This review did not verify primary evidence supporting those assertions. They have been removed, along with blanket assumptions that coupon partners are overpaid and earlier content touches necessarily created the purchase.
Plan a causal question separately
If the business needs to know whether a partner produces additional sales, write that question down before choosing a method. A qualified measurement reviewer can assess whether a suitable comparison, sufficient data and an interpretable result are feasible. Simply restricting a partner by device or geography does not automatically isolate its contribution. Other changes and differences between the groups can affect the result. Choose a method the team can explain before committing to a test.
The useful outcome may be a corrected report, an explained credit difference or a proposed study that needs further review. Each is progress when it answers the stated question. Avoid an automatic 10–15% commission bonus based only on appearing earlier in a recorded path. Compensation deserves a clear agreement and commercial reasoning, while a claim of budget waste deserves evidence of what was actually wasted.