No single Google-published decision tree exists for this; what follows is practitioner diagnostic methodology built from what’s actually observable, not an official Google process. The reliable approach is to check causes in order of how fast they can produce a sudden ranking drop, since the three causes have different characteristic timing signatures. SERP feature changes and intent shifts tend to produce sharp, dateable drops tied to a specific SERP composition change. Content quality erosion tends to produce a gradual, multi-week decline with no single inflection point. Checking in that order, fastest-changing cause first, keeps you from misdiagnosing a feature or intent problem as a content problem and rewriting a page that was never the actual issue.
Diagnostic sequence
Step one: check for a SERP feature change coincident with the decline. Pull historical SERP snapshots for the query, either from a rank tracker that stores SERP composition over time or from a web archive/screenshot tool, and look for the date the ranking dropped. Compare the SERP layout immediately before and after that date. Specifically check for: a new AI Overview or AI-generated summary appearing above organic results, a People Also Ask block expanding and pushing organic results down, a new rich result or vertical (video carousel, shopping units, a “things to know” panel) claiming space that previously went to organic listings, or a competitor’s content earning a featured snippet that it didn’t hold before. If a feature appeared right at the decline’s inflection point, and the page’s actual position within the remaining organic results hasn’t meaningfully changed, the decline is a visibility/feature problem, not a ranking-quality problem, and the fix path is different (targeting the feature itself, like structuring content to be snippet-eligible or AI-Overview-citable, rather than rewriting the page as if it lost quality ground).
Step two: check whether the query’s dominant intent has shifted. With the feature-change explanation ruled out or addressed, look at what kind of content now occupies the top of the organic results, independent of any special features. If a query that used to be answered by long-form informational content is now dominated by product/comparison pages, or a query that used to surface single-answer pages is now dominated by listicles or forum threads (a common recent pattern, where Google has increasingly surfaced Reddit and community-forum content for queries where searchers seem to want first-hand, unpolished opinions), that’s evidence the intent Google’s systems infer for that query has moved, and the page is now being evaluated against a different implicit standard than when it was originally optimized. This is checkable directly: compare the content type, format, and depth of the current top 5 against what the page itself offers, and against what the top 5 looked like before the decline (via the same historical snapshot tools used in step one). A page written to answer a research-stage informational query will structurally lose ground if the query has drifted toward a comparison-shopping or transactional intent, regardless of that page’s writing quality.
Step three: only if the first two are ruled out, investigate content quality. The distinguishing signature here is shape, not cause: quality-driven declines are typically gradual, a slow slide over several weeks or months as competing pages get updated, expanded, or refreshed, rather than a cliff on a single date. If the decline doesn’t line up with any SERP feature or intent-composition change from steps one and two, check whether competing pages in the same result set have been meaningfully updated or expanded since the page in question was last touched, whether the page has factual or informational elements that have gone stale (pricing, dates, procedures, statistics), and whether newer entrants in the SERP demonstrate more specific expertise or more current information than the page currently offers. This is the slowest and most labor-intensive category to fix because it requires an honest content audit against current competitors, not a quick structural correction.
Practical implication
Treat this as sequential filtering, not three simultaneous checks. Doing step three first, jumping straight to “the content must need improvement,” is the most common misdiagnosis in practice, because content audits always find something to improve whether or not that’s what actually caused the decline, and a rewrite undertaken to fix a feature or intent problem usually does nothing to the ranking, since the actual cause (a new AI Overview claiming the click, or the query now favoring a different content format entirely) is untouched by better prose on the same page in the same format.
Build a simple before/after SERP comparison as the first move on any decline investigation: date of drop, SERP screenshot or snapshot from just before, SERP composition just after, dominant content type in the top 5 before versus after. That single artifact usually resolves steps one and two immediately and either closes the investigation or correctly routes it to the slower content-quality audit in step three. Where the cause is a SERP feature, the response is usually structural (schema markup, snippet-optimized formatting, direct-answer placement) rather than a full rewrite. Where the cause is intent drift, the response may require a different content format entirely, not an improved version of the old one. Only genuine, gradual, competitor-outpacing decline warrants the deeper content-quality rework, and misapplying that response to the other two causes wastes the investigation and leaves the actual cause unaddressed.
Hypothetically, suppose a team notices a sharp, dateable ranking drop for “best budget espresso machines” on a coffee-equipment site. Pulling historical SERP snapshots might show an AI Overview appeared directly above organic results on the same date the ranking fell, with the page’s actual organic position essentially unchanged underneath it. Concluding the page needed a content rewrite would likely waste weeks, since the real cause was a visibility problem, not a quality one; the more productive response would be restructuring the page’s top section to be more directly citable in that kind of summary. By contrast, if the same team found no feature change but noticed the top results had gradually shifted from single-page buying guides to community forum threads over several months, that would point to an intent shift requiring a different content format, not a better-written version of the existing guide.
