Define the question first
Choose whether you are investigating a release, a traffic change, a content cluster or an indexation concern. The right date range, filters and comparison depend on that question.
Segment brand and non-brand demand
Brand queries can hide movement in discovery terms. Build stable definitions for brand, service, market and device segments, then record their limits instead of presenting them as perfect classification.
Read queries and pages together
A query can move because Google changed the ranking URL, several pages compete or demand changed. Compare query groups, landing pages, countries, devices and search appearance before deciding on a cause.
Account for data limits
Search Console rounds, samples and withholds some query data. Average position is an aggregate, not a rank tracker. Use the interface and API as directional evidence alongside crawls, analytics and release history.
End with a testable action
Record the suspected cause, supporting evidence, alternative explanations, action and measure that would confirm or reject the diagnosis. This turns a chart review into an operating decision.
Need this applied to your site?
Bring us the evidence, constraints and search problem. We will identify the useful next step.
Request an SEO review