New Google Discover Data Reveals Local-National Shakeup - And Why YouTube Quietly Won

Executive Snapshot

• Local news publishers largely retained in-state Discover presence but lost national exposure; Syracuse.com saw article placements drop 36% and its aggregated audience score fall 80%, with losses concentrated in Florida and California rather than New York [S1].
• Large brands such as Yahoo (≈50% fewer article placements, -62% audience-score rank from #3 to #9), Fox News/Fox Business/Fox Weather (>40% visibility losses), Forbes (-21% placements, -67% audience score), and X/Twitter (-22% placements, -32% audience score) all declined [S1].
• YouTube increased its Discover placements by 15% in the post-update window (16,283 → 18,803 placements), reinforcing the strength of Google-owned properties in this surface [S1].
• DiscoverSnoop and NewzDash show conflicting results for specific domains (for example, X.com and Geediting.com), driven mainly by different measurement windows and methodologies [S1].
• Some mid-tier publishers (Parade.com, Axios, Fortune, Newsweek, Wall Street Journal) show notable Discover gains, with Parade.com up 208% in placements and 1,300% in audience score [S1].

Implication for marketers: Treat Discover as a volatile, geo-sensitive traffic source, and rely on your own regional data rather than any single third-party view.

Google Discover core update - method and source notes

The figures here come from a Search Engine Journal summary of data from DiscoverSnoop (a commercial Discover analytics platform), particularly the DiscoverSnoop report, and earlier findings from NewzDash, plus Google’s own communication about the February Discover core update [S1]. These are not Google-published metrics, and there is no public raw dataset.

What was measured [S1]

• Timeframes
  • DiscoverSnoop: week before update start (26 Jan-1 Feb) vs week after update completion (2-8 Mar).
  • NewzDash: a mid-rollout window (dates not fully specified in the summary).
• Metrics
  • Article placements: how often a publisher’s content appears in Discover feeds monitored by the tool.
  • Audience score: a proprietary index combining placements and estimated audience reach (exact formula not disclosed).
  • Some state-level segmentation for selected local publishers (for example, New York vs Florida vs California).
• Scope
  • English-language Discover in the United States.
  • Emphasis on news and publisher-type sites (national brands, regional publishers, and some lifestyle sites).

Key limitations and caveats [S1]

• Vendor data, not first-party Google data. DiscoverSnoop and NewzDash capture only a subset of Discover feeds; their coverage and sampling strategy are not fully public.
• Short comparison windows. One-week pre- and post-periods capture immediate effects but not longer-term stabilization.
• Proprietary scoring. Audience score is relative within each tool; scores are not comparable across tools.
• Rollout timing. NewzDash and DiscoverSnoop measure different update phases (mid-rollout vs post-completion), which can materially change perceived winners and losers.
• No direct access to original datasets in this analysis. All figures are taken from the Search Engine Journal article summarizing the vendor reports and Google’s stated intent for the update [S1].

Findings from the Google Discover core update on local vs national publisher reach

The available data suggests a structural rebalancing of Discover visibility, with local relevance and Google-owned properties gaining ground, while some high-profile brands experience sustained declines [S1].

Large national brands with sustained Discover losses [S1]

• Yahoo
  • Nearly 50% reduction in article placements.
  • Audience score down 62%, dropping from 3rd to 9th in DiscoverSnoop’s ranking.
  • Discover visibility had already been declining since at least September; the core update accelerated this pattern.
• Fox properties
  • Fox News, Fox Business, and Fox Weather each saw visibility fall by more than 40%.
  • These domains were also trending downward in Discover prior to the core update.
• Forbes
  • Article placements: -21%.
  • Audience score: -67%.
  • Pre-update decline already in motion.
• X/Twitter (X.com)
  • Article placements: -22%.
  • Audience score: -32%.
  • This conflicts with NewzDash’s mid-rollout observation that posts from institutional X.com accounts rose into the Discover top 100 during that phase [S1].

Local publishers losing national reach but holding in-state visibility [S1]

• Syracuse.com (New York)
  • Overall: -36% in article placements, -80% in audience score.
  • State breakdown: relatively stable in New York; sharp declines in out-of-state feeds such as Florida and California.
• cbs6albany.com (New York)
  • Exhibited a similar pattern: home-state visibility more stable, with losses mainly in other states.
• Reinforcing evidence from NewzDash
  • Earlier analysis found New York-local domains appearing roughly 5× more often in the New York Discover feed than in the California feed after the update [S1].

Google said the update would show “more locally relevant content from websites based in their country” [S1]. The state-level patterns suggest the change extends beyond country-level adjustments toward sharper regional segmentation inside the United States.

Platforms and publishers that gained Discover placements [S1]

• YouTube
  • Article (video) placements increased 15% post-update, from 16,283 to 18,803.
  • The report notes that Google-owned properties rarely decline in Google core updates.
• Parade.com
  • Article placements: +208%.
  • Audience score: +1,300%.
• Other publishers with reported gains
  • Axios, Fortune, Newsweek, and the Wall Street Journal all show improvements in Discover placements and/or audience score.
  • Exact percentages for each outlet beyond Parade.com are not detailed in the summary [S1].

Edge cases and unexpected winners [S1]

• Geediting.com
  • DiscoverSnoop: article placements up 531%; audience score up 900% in the post-completion window.
  • More than 75% of titles reportedly start with “Psychology says…”.
  • DiscoverSnoop describes the site as not clearly aligned with the E-E-A-T profile (experience, expertise, authoritativeness, trustworthiness) that Google says it favors.
  • NewzDash: in its mid-rollout snapshot, a Geediting listicle slid from about #14 to around #153 in the observed feed.

These discrepancies highlight that update effects can reverse within the rollout cycle and that different tools may sample distinct segments of Discover.

Strategic implications of the Google Discover core update for marketers

This section interprets the data and labels confidence levels.

Likely: Discover distribution is now more region-weighted for local publishers

• State-level evidence for Syracuse.com and cbs6albany.com, plus NewzDash’s finding that New York domains appear about 5× more frequently in New York than California feeds, point to a system that amplifies local publishers mainly within their home region [S1].
• For local and regional outlets, this means Discover traffic numbers may skew more heavily toward in-state audiences, with reduced bonus exposure in distant states.
• Multi-state or national advertisers who relied on a single local brand’s Discover reach to cover multiple regions may see reduced scale and should not assume previous cross-state reach holds.

Likely: Large brand presence in Discover is no longer assured

• Yahoo, Fox properties, Forbes, and X/Twitter all show double-digit percentage declines in placements and audience scores, and in several cases the declines pre-date the core update [S1].
• Discover appears less dominated by a small group of large national outlets than it was previously, at least within the tracked sample.
• This change increases volatility for even well-known brands and reduces the reliability of Discover as a steady referral source compared with Search or direct traffic.

Likely: Google-owned surfaces retain strong priority

• YouTube’s 15% increase in placements during the same period that many news publishers declined reinforces the long-standing tendency for Google properties to maintain or grow visibility during algorithm shifts [S1].
• For brands already investing in YouTube, Discover may act as an additional distribution layer for video rather than a standalone strategy.

Tentative: E-E-A-T enforcement in Discover is inconsistent in the short term

• The rise of Geediting.com, characterized by repetitive “Psychology says…” headlines and described as misaligned with E-E-A-T expectations, suggests that Discover’s quality signals are imperfect or that engagement-heavy but thin content can still gain temporary traction [S1].
• Over a longer horizon, Google often iterates on quality updates, so the sustainability of such gains is uncertain.
• Marketers focused on durable growth are still better served by aligning with E-E-A-T principles rather than copying short-term outliers, but the presence of these cases explains why some low-trust content may appear in user feeds during and after core updates.

Tentative: Third-party Discover metrics should be used as directional guidance only

• Direct contradictions between DiscoverSnoop and NewzDash for X.com and Geediting.com, driven primarily by different measurement windows, show that any one snapshot can misrepresent a site’s true trajectory [S1].
• The proprietary nature of audience scores further limits comparability.
• Marketers can use these tools to track their own relative changes over time, but cross-vendor comparisons or precise market-share estimates are not reliable.

Tentative: Geographic segmentation is now essential for understanding Discover performance

• The pattern where local publishers retain home-state reach but lose out-of-state placements means aggregate Discover traffic figures can mask important regional shifts [S1].
• Business owners advertising in multiple states or running regional editions of the same brand benefit from tracking Discover traffic by state or DMA rather than only at domain level.

Speculative: Content types that signal localness may enjoy preferential treatment in local feeds

• Google’s intent to show more locally relevant content from publishers in a user’s country, plus the observed concentration of local publishers within their own states, hints that signals such as publisher location, coverage area, and possibly user-location interaction with content may have increased weight in Discover [S1].
• It remains unclear whether adding local context or signals to content (for example, clear geography in titles and metadata) materially changes Discover visibility, since the update description focuses on publisher origin rather than page-level cues.

Conflicting data and remaining gaps in Google Discover core update analysis

Several unresolved issues limit how precisely marketers can interpret these results.

Conflicting views on specific sites [S1]

• X.com
  • DiscoverSnoop reports a decline (-22% placements, -32% audience score) post-update.
  • NewzDash observed institutional X.com posts rising in prominence during the mid-rollout top-100 feed snapshot.
• Geediting.com
  • DiscoverSnoop shows triple-digit percentage growth; NewzDash recorded a drop in rank for one of its listicles during the rollout.

These contrasts illustrate timing sensitivity - a site can gain in one phase and lose in another - and tool sampling differences, since each vendor tracks different user cohorts and feed instances.

Short-term scope and lack of longitudinal data [S1]

• Both vendor snapshots focus on narrow one-week windows (pre-update, mid-rollout, post-completion).
• Core updates often include subsequent fine-tuning, which may moderate early extremes.
• Without data covering several months after the update, it is not possible to say whether observed gains or losses stabilized, intensified, or reversed.

Opaque metrics and coverage [S1]

• Audience score is not defined in a way that allows comparison across tools or with internal analytics such as Google Analytics or Search Console.
• The exact number and profile of Discover feeds monitored by each vendor are not detailed, so coverage bias (for example, focusing on particular devices, logged-in user segments, or geographies) may affect outcomes.

Unclear effects outside US English and beyond news-heavy sites [S1]

• The February update is limited to English-language users in the United States, with future expansion planned but not scheduled.
• No data is reported for non-news verticals (for example, niche blogs, product-focused content, or B2B publishers), so transferability of these findings to other categories is uncertain.

Marketers should treat all current figures as early indicators of direction rather than stable performance baselines.

Data appendix for Google Discover core update publisher performance

All numbers below reflect DiscoverSnoop and NewzDash data as summarized by Search Engine Journal [S1].

Pre- vs post-update changes (selected publishers/platforms)

• Yahoo
  • Article placements: ≈-50%.
  • Audience score: -62% (rank dropped from #3 to #9).
• Fox News / Fox Business / Fox Weather
  • Visibility: >-40% across these properties.
• Forbes
  • Article placements: -21%.
  • Audience score: -67%.
• X/Twitter (X.com)
  • Article placements: -22%.
  • Audience score: -32%.
• Syracuse.com
  • Article placements: -36%.
  • Audience score: -80%.
  • Geographic detail: New York state relatively stable; major losses in Florida and California.
• cbs6albany.com
  • Similar state-level pattern to Syracuse.com (home-state stability, out-of-state decline).
• YouTube
  • Article (video) placements: +15% (16,283 → 18,803).
• Parade.com
  • Article placements: +208%.
  • Audience score: +1,300%.
• Geediting.com
  • Article placements: +531%.
  • Audience score: +900%.
  • Title pattern: >75% of articles reported to begin with “Psychology says…”.

Source

[S1] Matt G. Southern, “Google Discover Core Update Data: Local Publishers Lost Reach,” Search Engine Journal, summarizing DiscoverSnoop and NewzDash tracking data and Google’s February Discover core update communication (article text supplied in prompt).