Google February 2026 Discover Update: 21 Days of Rollout in 2026
Google began the February 2026 Discover update on 5 February and took 21 days and 17 hours. Here is what it means for Discover traffic and how to respond.
On 5 February 2026, Google began rolling out an update affecting Discover, and it did not finish until twenty-one days and seventeen hours later. That is the longest rollout Google recorded in 2026 by a wide margin, roughly ten times the length of the fastest, and it deserves more attention than it generally received.
Discover is also the surface most publishers understand least, partly because it behaves nothing like search. There are no keywords to track and no queries to rank for, so most of the diagnostic habits people rely on simply do not apply.
What makes Discover different from search?
Search is a pull medium. Someone types something, and Google decides which pages best answer it. Discover is a push medium. Nobody asked anything; Google decides what to place in front of a user based on their interests and behaviour, and your content either gets surfaced or it does not.
That difference has an important consequence: Discover traffic is inherently volatile. A single article being picked up can produce a spike far larger than anything your search traffic does, and its disappearance can look catastrophic in analytics while nothing about your site has changed. Before concluding that an update harmed you, it is worth knowing that dramatic swings are the normal condition of Discover rather than a symptom.
Why a three-week rollout is unusual
Twenty-one days and seventeen hours is a very long deployment. Google typically rolls changes out gradually so that effects can be observed and moderated, and a period this extended suggests a change significant enough to warrant unusual caution.
What it means for you practically is that any analysis conducted during February was near-worthless. If you looked at Discover traffic on 12 February and drew conclusions, you were reading a system that had another two weeks of deployment ahead of it. The only comparison worth making is a settled period from March onward against a stable baseline from before 5 February.
How do you tell if a Discover change actually affected you?
Search Console has a dedicated Discover report, and it is the only reliable source. Ordinary analytics will show you a traffic drop but will not tell you whether it came from Discover, from search, or from somewhere else entirely, which is how a Discover decline routinely gets misdiagnosed as a search problem and treated with the wrong remedy.
Open that report and look at impressions rather than clicks first. If impressions fell, Google is surfacing your content to fewer people, which is a distribution change. If impressions held but clicks fell, your content is still being shown and something about its presentation, the headline or the image, is no longer earning the tap. Those are entirely different problems and confusing them wastes a lot of effort.
It is also worth checking whether the decline is spread across your catalogue or concentrated in particular topics. Discover is interest-driven, so a shift in which subjects get surfaced is common and does not necessarily indicate anything wrong with your site.
What actually influences Discover performance
Google's guidance for Discover is less prescriptive than its search documentation, but a few things are consistent. Content needs to be genuinely interesting to a person browsing rather than searching, which favours timeliness and genuine insight over comprehensive reference material. Large, high-quality images matter considerably more than they do in search, because Discover is a visual, card-based surface. Headlines need to be accurate rather than sensational, and clickbait is explicitly identified as something Google works to suppress.
The harder truth is that Discover cannot be engineered the way search can. There is no query to target and no position to optimise toward. What you can do is make your content genuinely worth surfacing and make sure nothing technical is preventing it from being surfaced at all.
The technical half nobody checks
That second part is where sites quietly lose. Discover eligibility depends on Google being able to crawl your content properly, on your images being large enough and accessible, and on your pages being structured clearly enough to be understood. A site with a broken sitemap, missing structured data, or crawler access problems can lose Discover visibility without any change in editorial quality, and no amount of better writing will fix it.
This is worth verifying rather than assuming, and audit.mohibraiskhan.com makes it a fifteen-second job with no signup needed to see your results. It checks more than forty technical factors across SEO, GEO and AEO, including your schema markup, your Core Web Vitals, whether your sitemap is genuinely reachable by a crawler, and whether AI crawlers are being blocked from your content, which matters increasingly as discovery shifts toward AI-generated answers as well as Google's own feeds. If the foundations are broken, that is the first thing to fix, and it is the cheapest thing to check. Run your free audit.
Frequently asked questions
How long did the February 2026 Discover update take?
Twenty-one days and seventeen hours. It began on 5 February 2026 and completed at the end of the month, making it the longest confirmed Google rollout of 2026.
Where can I see whether my Discover traffic changed?
Use the Discover report in Google Search Console, which appears for sites that have received Discover traffic. It is the only source that separates Discover performance from search performance, and looking at overall analytics instead is the most common reason people misdiagnose the cause.
Does this update affect my normal Google Search rankings?
It was reported as a Discover update, so the expected impact is on the Discover feed rather than on search results. If your search traffic changed during the same period, it is worth looking for a separate explanation rather than assuming a single cause.
Why is my Discover traffic so unpredictable even without an update?
Because Discover is a recommendation surface, not a query-response one. Whether your content is shown depends on user interests and behaviour, and a single article being picked up or dropped can move your totals dramatically. Large swings are normal and are not, on their own, evidence of a penalty.
Can I opt back into Discover if I lose visibility?
There is no opt-in or submission process. Eligibility comes from meeting Google's content policies and being technically crawlable and indexable. The practical route back is to fix technical barriers, use large high-quality images, avoid misleading headlines, and publish material genuinely worth recommending.
Audit