MONTH-END REPORT SIGN-OFF

Approve the report.
Not the ambiguity.

Run a five-part check before a recurring report reaches the client. The worksheet stays in this browser and produces a ranked risk plus a copyable acceptance note.

1 / CONTROL TOTALS

Can the output be reconciled?

$4,800 = $4,800

Synthetic example: report spend agrees with the independent platform export. A $100 difference would remain unresolved until corrected or explained.

Report totals agree with an independent source export.Compare spend, conversions and other decision metrics—not just row counts.

2 / METRIC DEFINITIONS

Would two reviewers calculate the same number?

30 ÷ 1,000 = 3%

Synthetic example: label the numerator and denominator. “Conversion rate” is ambiguous if one person uses platform conversions and another uses CRM-qualified leads.

Metric definitions name their numerator, denominator and source.The definitions approver is known and available to sign off.

3 / DATE LOGIC

Do the periods use the same clock?

Sep 1–30 vs Aug 1–31

Synthetic example: document timezone, inclusion rules and comparison period. Partial-month comparisons should be labeled rather than silently presented as complete.

Date windows, timezone and prior-period logic are documented.Late-arriving conversions and attribution lookback are handled consistently.

4 / CLIENT CLAIMS

Does the narrative stay inside the evidence?

Reported ≠ incremental

Synthetic example: platform-reported conversions may overlap. They do not by themselves prove unique leads, revenue, ROAS or causal impact.

Every client-facing claim is supported and limitations remain visible.Recommendations distinguish observed performance from attribution assumptions.

5 / ACCEPTANCE OWNER

Who can say the patch is done?

One person can approve metric definitions and the final report.Acceptance is based on documented checks, not a vague “looks right.”Recurring report typeDue within two weeks?Definitions approver

RUN THE CHECK

Rank unresolved handoff risk.

Check every statement that is already true. The worksheet turns the gaps into a plain acceptance note. No form is submitted and nothing is uploaded.

Use generic business context only. Do not enter client names, records, credentials or sensitive data.

BROWSER-LOCAL BY DESIGN

This worksheet does not upload or store your answers. If you choose to continue, the separate enquiry form explains what is stored and asks for consent. ReportPatch begins with redacted aggregate inputs and agrees secure handling before any review.

KEEP THE EVIDENCE WITH THE DECISION

A record someone else can verify

Record the frozen report version, source controls, metric-definition version, exact reporting window and time zone, refresh cutoff, output tests, unresolved risks, decision, approver, and decision time. A checked box alone is not evidence.

Synthetic record RP-AC-01	Value
Report and sources	R2; S1/S2 from RP-REC-01
Reconciliation	Source 1,000 USD; corrected report 1,000 USD; difference 0
Window	2026-09-01T00:00Z inclusive to 2026-10-01T00:00Z exclusive
Refresh and export tests	Frozen fixture recomputed; output total and expected sections checked
Unresolved risks	Upstream collection and causal attribution not tested
Decision and approver	Accept this limited fixture only; fictional analytics lead; 2026-10-27

A pass establishes the agreed checks for the specified snapshot. It does not validate every source event, future refresh, or campaign claim. Carry unresolved risks into accept, revise, or hold rather than silently clearing them.

Download the acceptance record and template · Run the export checks · Inspect the synthetic evidence pack