The February 2026 Discover update, US English first, targets clickbait and rewards topic-level expertise. How did your Discover traffic shift by section?

1 reply

• 

  adminAdmin

  Original poster 1 hour ago

  There is no clean, public section-by-section dataset for this update that I know of, and I would be cautious about treating any single site’s before and after as the answer. What can be said with more confidence is where the shifts should show up if the update works as Google described it: unevenly by topic, not evenly across a domain. Sections with a sustained body of original work on one subject are the ones positioned to hold or gain. Sections built on occasional trend-chasing outside the site’s core, on sensational headlines, or on content aimed at US readers from outside the US are the most exposed.

  That also makes a sitewide Discover total the wrong number to judge. A flat total can hide one section gaining and another collapsing. The useful work is measuring per section (ideally per topic), normalised for how much each section published, and using non-US traffic as a rough control for the period when the rollout was US-only.

  What Google actually changed, and what it left unsaid

  Google announced the February 2026 Discover core update on 5 February, and it finished rolling out on 27 February. It applied first to English-language users in the United States, with Google saying it would expand to all countries and languages in the coming months. The stated changes were three: more locally relevant content from sites based in the user’s country, less sensational and clickbait content, and more in-depth, original and timely content from sites with expertise in the topic. Search Engine Land noted that non-US sites publishing for US audiences could see reductions until the wider rollout.

  The most useful line for this question is Google’s example of topic-level expertise, as reported by Search Engine Land: “A local news site with a dedicated gardening section could have established expertise in gardening, even though it covers other topics. In contrast, a movie review site that wrote a single article about gardening would likely not.”

  What Google did not say matters as much. It did not define a “topic”, did not say that URL folders or site sections map to topics, and did not say how much coverage a site needs before it counts as having expertise. The Discover documentation still carries the familiar guidance (avoid clickbait and “sensationalism tactics that manipulate appeal by catering to morbid curiosity, titillation, or outrage”) and still warns that Discover traffic is “less predictable or dependable” than keyword-driven search visits.

  Check before comparing markets:

  I could not confirm whether the promised expansion has reached every country and language yet. Before attributing a non-US change to this update, check the Google Search Status Dashboard and Google’s announcements for your market.

  Why the section is the right unit, with one caveat

  Think of how a bookshop buyer decides stock. The shop keeps reordering a publisher’s cookery list because readers trust it, and that decision says nothing about the same publisher’s one-off travel title. Google’s description reads the same way: credibility is earned per subject, not granted to the whole masthead.

  The caveat is that your CMS sections are a proxy. Google infers topics from content, not from your URL structure. A /lifestyle/ folder might hold gardening, relationships and celebrity news, and those could be treated very differently. Where a folder mixes topics, segment by your topic tags or categories instead, even though it takes more work.

  What the public data showed

  The main public dataset is NewzDash’s tracking of Discover feeds, summarised by Search Engine Journal:

  • Unique domains in the US top 1,000 Discover placements fell from 172 to 158, and California showed a similar contraction. Fewer sites were winning top placements.
  • Topic variety increased while publisher diversity decreased: more subjects, served by a narrower set of publishers.
  • New York local domains appeared roughly five times more often in New York’s feed than in California’s.
  • Posts from institutional X accounts in the US top 100 rose from 3 to 13, continuing a trend that started before the update.

  Treat these as signals about how placements concentrated, not as traffic data. Tracked feeds are samples, not your audience’s personalised feeds, and placement counts are not clicks. They are consistent with the “fewer, more topic-specific winners” reading, but they do not tell you which of your sections moved.

  Where I would expect gains and losses by section type

  These are hypotheses built from Google’s stated criteria, not observed results. The last two columns are what you should look for in your own data.

  Section type	Exposure under the stated criteria	Result that would support it	Result that would contradict it
  Core beat with a deep archive and named specialists	Best placed to hold or gain	Stable or rising impressions per article in the US after the update	Losses concentrated here, which would point to images, headlines or technical issues instead
  Occasional off-topic trend pieces	Most exposed: weak topic history and often sensational framing	A falling share of new articles reaching any Discover impressions	An unchanged share in the US after February
  Curiosity-gap headlines inside strong sections	Exposed at article level even where the section is strong	A widening gap between headline styles within the same section	No difference between headline styles
  Non-US site or desk writing for US readers	Exposed on local relevance	US Discover down while home-market Discover holds	Both markets down equally, which suggests quality rather than locality
  Syndicated or lightly rewritten wire copy	Probably exposed on originality, less certain	Original pieces in the same section pulling further ahead of rewrites	No change in the gap between original and rewritten pieces

  Confounders that can masquerade as a section effect

  Discover changed a lot around this update, so a section’s drop may have nothing to do with how Google judged your expertise.

  • More competition for the same slots. Google added posts from X, Instagram and YouTube Shorts to Discover, plus a follow feature, from September 2025. The NewzDash X figures suggest that trend kept growing.
  • Generative AI features in Discover. Google’s generative AI performance reports count impressions for them, but show impressions only and only reached all sites worldwide on 31 August 2026, so they cannot cover the February windows.
  • Follow signals. Google Search profiles (June 2026, US first) make followed content more likely to appear in Discover. This only matters if your post-update window runs into mid 2026.
  • A Search core update soon after. The March 2026 core update started on 27 March and finished on 8 April. That leaves a clean post-update window of roughly 6 to 26 March. Keep your comparison inside it so you are not blaming one update for two updates’ effects.
  • The news cycle and your own output. Sports seasons, elections, a quiet month for celebrity news, fewer articles published, image size changes or a missing max-image-preview:large setting can all move a section on their own.

  How to measure the section shift so it means something

  1. Export Discover data by page and date. The Search Console UI is fine for a handful of sections. For anything more, use the Search Analytics API with type=discover and the page dimension. There are no queries in Discover, so pages are your only lens.
  2. Map every URL to a section and a topic, and join publish date and a headline style tag from your CMS.
  3. Choose clean windows. Compare several weeks before 5 February with an equal period starting at least a week after 27 February (Google recommends waiting at least a week after a rollout completes), matched by weekday. Add the same weeks of 2025 to control for seasonality.
  4. Normalise for volume and hits. Discover is hit-driven, so totals mislead. Track impressions and clicks per article published, the share of new articles that received any Discover impressions, and the median article rather than the mean.
  5. Split by country. While the rollout was US English only, non-US Discover traffic on the same sections works as a rough control. A section that fell in the US but not elsewhere in February points towards the update. A section that fell everywhere points elsewhere. Audiences differ between markets, and the control disappears once the update reaches them, so treat it as supporting evidence, not proof.
  6. Separate impressions from CTR. Falling impressions suggest Discover is showing the section less. Stable impressions with falling CTR suggest a presentation problem (headline, image) or a change in which placements you now get.

  Section lost Discover traffic

  ↓

  Still down per article published? If not, it is a volume change, not a ranking change

  ↓

  Down in the US only, or in every market? Everywhere suggests a cause other than this update

  ↓

  Impressions down, or only CTR? CTR alone points to headlines and images

  ↓

  Core topic you intend to own, or an occasional traffic play? Decide whether to invest or accept the loss

  Hypothetical example:

  A US regional news site sees Discover clicks roughly halve in its entertainment section while local news rises and recipes dip slightly. Before blaming the update, the team finds entertainment also published far fewer articles in the post-update window. Per article, the drop is smaller but still real, and it is concentrated in curiosity-gap headlines about national celebrities, while entertainment pieces about local venues held. Non-US traffic to the same section barely moved. That combination fits the update’s stated criteria much better than a sitewide penalty, and it suggests the fix is editorial focus rather than technical work.

  What to do with a section that lost

  If the section covers a topic you want to be known for, the work is the slow kind: original reporting, named writers with a visible track record in that subject, consistent publishing, and headlines that describe the story. Expect movement over months, possibly only after a later update, not within days of rewriting titles.

  If the section was an occasional traffic play outside your core, check what the traffic was worth before trying to rescue it. Discover sessions from off-topic viral pieces often carry lower revenue per session than the core audience, and the editorial hours might earn more in sections where you already have standing. Accepting the loss can be the right commercial decision.

  For headlines, test rather than rewrite everything to be blander. Run straight descriptive headlines against your usual style within the same section and compare Discover CTR and impressions per article over several weeks.

  If this were my site, the one number I would take into the next editorial meeting is Discover impressions per article by topic, US against non-US, before and after February. Sections that hold on that measure deserve more investment. Sections that only ever did well through volume or curiosity-gap framing need a clear decision: build real depth in the topic, or stop counting on Discover for them.

  Need help with this on your own site?

  Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.

  See SEO services