How to confirm report scope and date window match the decision
A GA4 report can show movement that looks actionable while answering a question nobody asked. The reviewer should confirm the property scope, report name, date range, and comparison period match the exact decision under review. A weekly trend report cannot answer a quarterly budget question. A property scope that mixes staging and production data cannot support a live channel recommendation.
Timezone mismatches are easy to miss and hard to undo. If the property timezone does not match the business operating calendar, every daily comparison shifts by hours that nobody notices until a decision is already made. The reviewer should verify the timezone setting and confirm the owner knows which surface is being reviewed before accepting any number as evidence.
- Match report name, property scope, and date range to the decision
- Verify timezone setting against the business operating calendar
- A weekly report cannot answer a quarterly question
- Confirm the owner can identify the exact reviewed surface
How to check acquisition source and channel meaning before reading
First-user acquisition and session traffic acquisition answer different questions with the same channels. The first-user report shows where users originally came from. The session report shows how current visits arrived. Using one to make a decision that belongs to the other misreads channel performance completely. The reviewer should confirm which acquisition scope applies before interpreting movement.
Channel groupings are not neutral. GA4 default channel groups classify traffic by rules that may not match how the business thinks about its channels. A source-medium pair that lands in Display when the team considers it Paid Social produces a recommendation attached to the wrong channel. The reviewer should check grouping rules and verify that source-medium evidence matches the channel label in the recommendation.
- Separate first-user acquisition from session traffic acquisition
- Confirm channel grouping rules match business channel definitions
- Check source-medium pairs against their assigned channel label
- A display-classified source read as paid social misleads the decision
How to verify landing page and page identity before acting
Page identity is evidence, not formatting. A landing page report that shows not-set rows, query string fragmentation, or duplicate page titles cannot safely support a page-level recommendation. The reviewer should check whether the page path identifies a single surface the owner can locate. If the same page appears as five rows because of trailing slashes or UTM parameters, the identity is broken.
Page title mismatches create a separate risk. A report showing strong engagement on what appears to be a blog post may actually reflect a product page with a reused title. The reviewer should cross-check page path and page title before accepting any page as the subject of a recommendation. An action assigned to the wrong page wastes the work and erodes trust in the next GA4 read.
- Check for not-set rows and query string fragmentation in page reports
- Verify page path identifies a single surface the owner can locate
- Cross-check page title against page path before accepting identity
- Hold when page identity is ambiguous or split across multiple rows
How to separate engagement events from conversion evidence
Not every event that fires represents a business outcome. A scroll tracked as a conversion trains algorithms and inflates channel credit. The reviewer should classify every event as engagement signal, business action, or marked conversion before using it as decision evidence. An engagement event informs behavior. A conversion event steers budget. Using a scroll to justify a spend shift is a category error.
Conversion role requires proof beyond the event name. The reviewer should confirm the event is marked as a conversion, the parameters carry business value, and recent debug evidence proves the exact journey the report depends on. An event labeled purchase in the report but unverified in debug view is a label, not proof. Hold any recommendation built on conversions that have not been tested end to end.
- Classify every event as engagement, business action, or conversion
- Conversions steer budget. Engagement events inform but never decide spend
- Confirm conversions are marked, parameterized, and proven in debug view
- Hold recommendations built on conversions that lack end-to-end test proof
How to attach missing evidence to a named hold condition
A checklist item that fails is not noise. It is a hold condition with a name. The reviewer should attach every failed check to a named owner, a specific missing input, and the impact on the recommendation. A scope mismatch is not a preference. It means the report cannot answer the question. A page identity gap is not a formatting issue. It means the action may land on the wrong surface.
The approval state should match the evidence state. A checklist can approve a follow-up review without approving the downstream change. If evidence is missing, the output is a held recommendation with the gap named, not a weaker approval. The next action stays gated until the reviewer accepts the readiness read. Moving forward before tracking is validated turns the checklist into paperwork instead of a gate.
- Attach every failed check to a named owner and specific missing input
- State the impact on the recommendation for each gap
- Approve follow-up reviews separately from downstream changes
- Gate the next action until the reviewer accepts the readiness read
Sample Review Note
The reviewer confirms report scope, property, date range, comparison window, and timezone match the decision. Acquisition meaning is separated: first-user and session traffic are not mixed without a written caveat. Channel grouping rules are verified against business channel definitions. Landing page identity is confirmed: no not-set rows, no query string fragmentation, page path and title agree on a single surface the owner can locate.
Events are classified as engagement, business action, or conversion. Marked conversions have business parameters and end-to-end debug proof matching the journey the report depends on. Every failed checklist item is attached to a named owner, missing input, and impact statement. The approval state matches the evidence: follow-up reviews may be approved while downstream changes stay held. If any property scope, date range, channel grouping, page path, event definition, or conversion marker changes after this review, the checklist is gated for recheck.