When a client wants to present attributed sales as proven incremental growth, decline that wording and offer a supported replacement. Name what the data measures, keep the limitation beside the result, and ask which decision the client needs to make. Record the agreed wording before the deck leaves the agency. Client approval can authorize a presentation; it cannot supply missing evidence of causation.
Find the decision behind the requested claim
Start with two questions: who will see the claim, and what should they decide? An internal renewal discussion, a board presentation, and a public agency case study need different review paths. All need wording that matches the evidence.
A client asking for "proof of growth" may need to defend a budget rather than publish a causal estimate. You can help with that decision while declining an unsupported headline.
Try this response:
We can report the revenue credited to the creator campaign under the agreed tracking method. That result supports a discussion about renewal. We have not measured how much revenue would have occurred without the campaign, so we cannot describe the full amount as incremental growth. Is the decision to renew at the current budget or increase it?
This is an illustrative response, not a script taken from a client engagement. The review sheet and rewrites below are also illustrative operating templates.
Locate the change in meaning
Google Analytics defines attribution as assigning credit for user actions across touchpoints. A rule or algorithm determines that allocation. A report showing sales credited to a campaign answers a measurement question about that allocation.
A claim that the campaign caused additional sales asks what would have happened without it. For example, Google Ads Conversion Lift uses treatment and control groups to measure incremental conversions. That product has account-availability limits and does not establish lift for an unrelated organic creator campaign.
Do not dismiss every modeled result as meaningless. Google's data-driven attribution uses a modeled counterfactual approach. But that does not turn a campaign's attributed revenue total into a completed holdout study. Ask the analyst what the particular method estimates and which assumptions support the claim.
For this review, flag words such as "caused," "incremental," "proven," and "would not have happened." Each introduces a claim that needs evidence beyond a tracked-sales total. A sales increase after posting also needs review before anyone assigns the increase to creators.
Modash's reporting guide recommends documenting the attribution model and using consistent methods. Make that documentation part of the evidence review, while assessing causal claims separately.
Use one claim-evidence review sheet
Review the proposed sentence before editing the surrounding deck. The following completed example assumes a hypothetical campaign with $80,000 in attributed net sales. Every number and operating detail is invented for illustration.
| Review field | Illustrative entry |
|---|---|
| Requested wording | "Our creator campaign generated $80,000 in incremental sales." |
| Intended use | Internal quarterly budget presentation; no public reuse approved |
| Decision | Renew at the current budget or increase spend |
| Available evidence | Client-approved affiliate export credits $80,000 in net sales to the campaign for August 1-31 |
| Counting rule | Agreed 30-day click window; refunds through September 15 deducted; duplicate orders removed |
| Evidence missing | No accepted estimate of sales without the campaign |
| Supported wording | "The affiliate report credits $80,000 in net sales to the August creator campaign under the agreed attribution rules." |
| Required limit | "Incremental sales were not measured." |
| Recommended decision | Consider renewal using the agreed performance criteria; evaluate a measurement plan before claiming a causal return from increased spend |
| Owners | Agency analyst checks the export; account lead owns wording; client sponsor approves the decision and intended audience |
| Release condition | Analyst confirms the amount and rules; account lead replaces the causal claim before circulation |
Keep the export version, report dates, filters, revenue definition, and review date with the sheet. These details let another reviewer trace the sentence to its source. Do not add individual customer records to a presentation merely to make the evidence look stronger.
Choose one of three outcomes for each claim:
- Use: The evidence supports the sentence, its scope, and its stated level of certainty.
- Rewrite: The observed result is usable, but the requested interpretation exceeds it.
- Hold: The amount, method, access, or claimed effect cannot yet be checked.
These are recommended agency controls. They are not platform-defined approval statuses.
Resolve access gaps before debating adjectives
Sometimes the account team cannot inspect the report behind the client's number. Label it as client-supplied until an authorized reviewer can check the source and settings. Missing access is an unresolved question, not evidence that the campaign worked or failed.
For a Google Analytics workflow, ask the client's administrator for the least access needed. A property-level Viewer role may suit report inspection. Google's access documentation says Viewer can see settings and data, subject to restrictions. It also describes a No Revenue Metrics restriction. Restricted values can appear as zero in reports, so check permissions before interpreting zero revenue.
Have the administrator confirm effective permissions, including inherited restrictions. If direct access is unavailable, request an authorized export with its date range, filters, and settings, plus confirmation from the client analyst. Keep the supplied-report label. Do not ask colleagues to share passwords or change attribution settings to obtain a more favorable result.
More access can resolve a source-checking problem. It cannot retroactively create an incrementality study.
Three client-facing rewrites
The examples below use hypothetical figures. Each preserves a usable result while removing a claim the stated evidence cannot support.
1. Attributed sales presented as additional sales
Requested: "Creators added $80,000 in revenue that we would otherwise have lost."
Replacement: "The affiliate report attributes $80,000 in net sales to the creator campaign under the agreed rules. We have not measured how much of that revenue would have occurred without the campaign."
Use this after verifying the attributed total. Keep the limit in the same paragraph or slide as the result.
2. A monthly increase presented as campaign lift
Requested: "The campaign drove 20% sales growth."
Replacement: "Store sales were 20% higher during the campaign month than the previous month. This comparison does not isolate the creator campaign's contribution from other changes."
Use this only when the comparison is verified and relevant. Name material changes, such as a concurrent promotion, when known. If the periods are not comparable, hold the percentage rather than trying to repair it with a footnote.
3. Attributed efficiency presented as guaranteed profit
Requested: "Our 4x return proves we should double creator spend."
Replacement: "Attributed net sales were four times the recorded campaign spend. This ratio does not measure incremental profit or predict the return from doubling spend. We recommend evaluating a staged budget increase against agreed cost and measurement criteria."
Confirm what the spend total includes. If it omits agency fees or product costs, state those exclusions rather than calling it the full campaign cost.
Set the release boundary when the client insists
Send the original sentence, proposed replacement, and missing evidence to the client sponsor together. Ask for a decision on the replacement by the deck's release deadline. The analyst owns the evidence assessment; the account lead owns escalation. A commercial deadline should not leave a junior analyst responsible for accepting an unsupported claim.
If the client still requests the original, escalate to the agency's responsible director. Seek qualified legal review where the proposed use raises advertising or contractual questions. Hold the unsupported sentence while allowing supported sections to proceed where practical. Keep a written record of the agency's advice and the version it approved.
For US public advertising, the FTC's substantiation policy says advertisers and ad agencies need a reasonable basis for objective claims before dissemination. Claims that imply a particular level of proof also need that support. A public case study promoting agency services deserves that review. This policy does not by itself decide the legal treatment of a private board deck. Ask qualified counsel about the intended use and jurisdiction; this workflow is not legal advice.
Once wording is settled, use the campaign-report guide to organize the recommendation and evidence. If the client needs a causal answer before increasing spend, use the incrementality guide to scope the next measurement discussion.
Before the next client call, copy the disputed sentence into the review sheet. Put the supported replacement beside it, name the missing evidence, and assign one person to approve the release wording.
