Segment Search Console performance data by URL pattern, page type, or content category for the specific date window the core update rolled out, and compare the decline’s breadth across those segments. A decline concentrated in specific templates or content categories, thin category pages, low-effort programmatic listings, a particular content vertical, points to a page-type-specific quality issue in that segment. A decline spread broadly across most page types and templates on the domain points to a domain-level trust or quality reassessment rather than an isolated content problem.
Mechanism: why segmentation is the diagnostic, not aggregate traffic
Google’s own guidance on core updates is explicit that these updates reassess content more broadly, they’re not targeted at fixing specific individual pages or issues, but at reassessing how Google’s systems evaluate content relative to what’s available across the web. That reassessment can land unevenly across a site: some page types might reflect exactly the kind of comprehensive, trustworthy content Google’s updated systems reward, while other sections of the same domain, if they’re thin, templated, or don’t demonstrate genuine expertise, might not.
Aggregate, whole-site traffic data obscures this distinction. A 20% overall traffic drop could mean every page type lost roughly 20%, consistent with a domain-wide trust reassessment, or it could mean one page type (say, a large programmatic content section representing a third of total traffic) lost 60% while everything else held steady, which is a completely different diagnostic conclusion and requires a completely different fix. Only segmentation reveals which of these actually happened.
Practical segmentation steps
Filter Search Console Performance data by URL path pattern or page type, using the date comparison feature to isolate the specific rollout window Google announced for the core update (Google publishes start and end dates for core update rollouts, and the comparison should bracket that window precisely, not a vague “before and after” split that includes unrelated time periods).
Break the site into its natural content categories or templates (blog content vs. product pages vs. category/listing pages vs. evergreen guides, or whatever the site’s actual structural divisions are) and compare each segment’s performance change independently rather than relying on one blended number.
Check whether the decline correlates with a specific content characteristic shared across the affected segment, thin pages, heavily templated/programmatic pages with minimal unique content, content lacking clear authorship or expertise signals, older content that hasn’t been meaningfully updated, versus segments that don’t share that characteristic and were unaffected.
Confirm the pattern isn’t actually device-, geography-, or query-type-specific before concluding it’s a page-type pattern, since core updates can sometimes interact differently with different query intents even within what looks like one content category, and ruling out these confounds keeps the page-type conclusion honest.
What each pattern implies for next steps
If the decline is genuinely page-type-specific, the diagnostic conclusion points toward Google’s own self-assessment questions (from its “creating helpful, reliable, people-first content” guidance) applied specifically to the affected segment: does this content demonstrate real expertise and firsthand knowledge, does it exist primarily to rank for search terms rather than to serve a genuine reader need, would someone leaving this page feel they’d learned something substantial. A page-type-specific decline suggests the fix is concentrated content-quality work on that segment specifically, not a site-wide overhaul.
If the decline is broad and spans most page types and templates, that’s more consistent with a domain-level trust or overall-quality reassessment, meaning Google’s updated systems are evaluating the site’s overall pattern of trustworthiness and helpfulness differently, not flagging one weak section. This is a harder, slower problem: it typically requires honest self-assessment against the same helpful-content questions but applied to the site’s overall pattern and reputation, not a single segment, and Google has been explicit that there’s no quick technical fix for a broad quality reassessment; recovery (to whatever extent it occurs) tends to require sustained, demonstrable improvement across the site’s general content pattern, reflected over the following months and typically becoming visible around a subsequent core update rather than immediately.
A hypothetical walkthrough of the segmentation process
Imagine a hypothetical site, “Example Home Services Hub,” that sees an overall 20% traffic drop in the window bracketing a core update rollout. Hypothetically, if the team stopped at that aggregate number, they might conclude the whole domain lost Google’s trust and start a broad, unfocused overhaul. Now imagine they instead segment Search Console data by page type and find that the site’s evergreen how-to guides held essentially flat, while a large programmatic section of thin, templated city-service pages dropped sharply, enough on its own, given its share of total traffic, to produce that 20% blended figure. In this hypothetical, the segmentation would completely change the diagnosis: not a domain-wide reassessment, but a page-type-specific quality issue concentrated in one template. The fix that follows from that hypothetical finding, targeted content-quality work on the thin city-service template, is a different and far more tractable project than the sitewide response the aggregate number alone would have suggested.
Practical implication: don’t skip the segmentation step
The common diagnostic mistake is jumping straight to broad, site-wide explanations (“Google doesn’t like us anymore”) or, conversely, fixating on one visible weak page without checking whether the decline is actually broader than that one example. Segmentation first, before forming a theory about cause, keeps the diagnosis evidence-based rather than anecdotal, and it directly determines whether the appropriate response is a targeted content-quality intervention on a specific segment or a more fundamental, sustained reassessment of the site’s overall content approach.
