In early February 2026, Google began rolling out a specific core update for Google Discover, focused on English-speaking users in the United States.
The official announcement mentioned three objectives: more locally relevant content, less clickbait, and more in-depth and timely content from sites with thematic authority.
Sounds good on paper. But Google ads always sound good on paper.
The interesting thing is that this time there is data from third parties that allow us to verify if what Google said was going to happen really happened.
NewzDash, a platform of SEO for media, it compared Discover feeds before and after the update using panel data from millions of users in the United States, California and New York.
They compared the pre-update window (25-31 January) against the post-update window (8-14 February).
This is what the data shows.
Of the three objectives announced by Google, this is the one with the clearest evidence.
The California and New York feeds are not the same as the national feed. Each one has a local content layer that the other one doesn't. Local New York domains appear approximately five times more often in the New York feed than in the California feed, and vice versa.
In California, local articles in the top 100 of the feed went from 10 to 16 in the post-update window. Publications such as SFGate, LA Times and Sacramento Bee appear in positions they did not hold before.
In New York, the pattern is different, but the logic is the same: the feed includes content from NJ.com, Newsday, and media from the metropolitan area that don't appear in the national or California feeds.
For regional publishers, this is the best news from the update. The structural advantage of a local environment, which is to cover what no one else can cover with the same depth and relevance, now seems to have more algorithmic weight.
If you're a regional media outlet or your client is a local publisher, this update is a real opportunity.
For the rest of the marketing teams: If your audience is geographically concentrated, the local relevance signal matters more than before.
This is the most difficult objective to measure. “Clickbait” isn't a technical category, it's a combination of headline, thumbnail, implied promise, and page experience. It can't be detected just by looking at spreadsheets.
What can be detected is which specific publishers lost visibility and what type of content they were producing.
What these cases have in common is that the content followed systematically repeatable templates. It wasn't just an isolated curious headline, it was a production model based on curiosity gap patterns applied at scale.
The distinction that emerges from the data is important: engagement is not penalized. Headlines that generate interest continue to work when they are backed by real thematic authority.
What seems to be penalized is the template without substance behind it, the pattern applied en masse without differentiation.
For content teams: If your headline strategy depends on formulas such as “this is what no one tells you about X” or “the secret behind Y” applied systematically, the risk increased.
If your headlines are direct but your content has real depth, there's probably nothing to change.
Here is the most interesting tension of the update, and the one that should concern average publishers the most.
The number of unique categories in the feeds increased in all the geographies analyzed. The American feed went from 163 to 173 unique categories. California from 162 to 170. New York from 168 to 183. Discover is covering more thematic niches.
But at the same time, the number of unique domains in the top 1000 fell in the United States (from 172 to 158) and in California (from 187 to 177). Traffic is more concentrated.
What this means in practical terms: Google is choosing to cover more topics, but it's being more selective about which publishers represent each topic. It's not “everyone can participate in more categories.” It's “more categories, more curated publishers.”
The pattern is consistent with what Google described as a “topic-by-topic” evaluation: a site may not be the biggest or the most well-known, but if it has demonstrable depth and authority in a specific area, it can gain distribution in that niche. The generalization without depth is the one that loses.
New York shows a slightly different pattern, with unique domains that increased marginally (from 192 to 194) and concentration that fell slightly. But in the national aggregate, the direction is clear: consolidation.
This is the most unexpected fact of the analysis, and the one with the most unresolved implications.
X posts in the top 100 of the American feed went from 3 to 13 between the pre and post-update windows. In New York, from 2 to 14. In the standard domain analysis, the visibility of X in Discover rose 38%.
Most of the high-ranking X items come from institutional accounts: media, political figures, news outlets. These are not random personal accounts.
One possibility is that Google is treating posts from authoritative accounts in X as “timely and original content” within its Goal 3 logic. If that's the case, your presence on X begins to be part of your distribution surface in Discover, not just a parallel channel.
But there is a complication. If Discover sends the user to an X post, and from X the user reaches your article, that's an additional step along the way. That extra friction probably reduces click-through to your own pages for a percentage of users.
The NewzDash analysis explicitly mentions the NYT case: nytimes.com appears less often in the post-update top 1000 (8 to 5 articles), while an NYT post on X is the number 1 item in the post-update feed.
Whether that is causation or coincidence cannot be confirmed with the available data. But it's a hypothesis worth monitoring.
For publishers and marketing teams: If you're not measuring how much Discover traffic comes via X versus direct to your site, now's the time to implement it.
The mix of content that Discover shows at the top of the feed changed. News rose from 15.94% to 19.16% of weighted participation in the US feed. Sports from 5.06% to 8.52%. Arts and entertainment fell from 24.40% to 17.90%.
A clarification of context needs to be made: the post-update window coincides with the Super Bowl, the Winter Olympics and the T20 World Cup cricket. Some of the increase in sports is likely to reflect the news cycle, not just the algorithm.
But the drop in entertainment is consistent in the three geographies analyzed and there is no specific event that explains it in the same way. The feed is moving towards content that is more informative and moving away from content for passive consumption.
Three specific things to take away from this analysis.
Thematic authority now has more weight than volume. Covering a lot superficially is the strategy that loses with this update. Covering a specific area well wins. If your blog or your client's blog is trying to touch on every possible topic, now is the time to reconsider.
Local content has a real window of opportunity. If you have an audience focused on a specific geography, the local relevance signal matters more than before. That applies both to regional publishers and to brands with a strong local presence.
X is a new variable in the Discover equation. It's not urgent to change your entire strategy because of this, but it's worth starting to measure if your X posts appear on Discover and what happens to traffic when they do.
The update is still being rolled out. The NewzDash data covers an early window and some patterns may change. But the direction is clear: Discover is being more selective, more local, and more oriented to real utility than to entertainment consumption.
.png){kind=small}
Consultoría SEO personalizada para escalar tráfico, posicionar tu marca, aumentar el revenue y dominar el mindshare en Google y modelos de IA como ChatGPT.
%402x.svg){kind=icon}
