Search in 2026 is less predictable for teams still working from the 2020 playbook. Targeting a keyword, matching intent, earning links and securing a high position can still matter, but none of those things guarantees the traffic outcome it once appeared to.
The clearest evidence is not one catastrophic algorithm change. It is the frequency with which publishers and SEO teams now have to separate ranking, impressions, click-through rate and actual visits instead of assuming those metrics will move together.
Google’s confirmed update history explains part of that instability. The March 2026 core update began on March 27 after the December 2025 core update and a February Discover-only update. Google then released a separate May 2026 core update, which ran from May 21 to June 2.
The June 2026 spam update followed later that month and completed in roughly two days. The March spam update before it had finished in under 20 hours, making it the fastest confirmed spam rollout in Google’s dashboard history. The August 2026 spam update then ran from August 18 to August 21.
Between those named events, third-party volatility tracking repeatedly showed sharp movement. Search Engine Roundtable flagged another spike on January 21, reported that the heated conditions continued later that month, and found only partial signs of cooling in February.
The distinction matters. Those tracker movements are not all confirmed Google updates, and they should not be presented as if Google had announced them. They do show why practitioners increasingly experience ranking volatility as an operating condition rather than a once-a-quarter event.
Barry Schwartz, who has covered search marketing for more than two decades, described the change directly in March. In a March 2 volatility report, he wrote that the movement had become frequent enough that he had reached the point of covering it weekly rather than posting about it every day.
What changed underneath the results page is a different question, and there is no single public Google document that explains every ranking movement. What can be described more confidently is how the search experience itself has changed.
One of the clearest shifts is the growing distance between being visible and earning the click. When Google displays an AI-generated answer above the traditional results, a publisher can continue appearing in the search experience without receiving the same proportion of visits.
Data reported by Search Engine Journal from Seer Interactive illustrates the pattern. In Seer’s dataset, brand-cited pages in AI Overviews reached 39.5 million impressions in November while clicks fell to 301,783 and click-through rate dropped to 0.76%. Earlier in the period, impressions had also grown much faster than clicks.
That does not prove every publisher will see the same curve, and it does not mean every traffic decline can be blamed on AI Overviews. It does establish that visibility and traffic can decouple in AI-heavy search results. A page can continue appearing while delivering fewer visits than its impression count might suggest.
That is particularly relevant to informational content. Definitions, simple how-to articles and pages designed primarily to answer a clean factual question now compete not only with other publishers but with the answer Google places directly on the results page.
The second shift is less settled: the industry’s growing focus on brand and entity signals. Practitioners increasingly talk about branded search demand, direct audiences, mentions beyond Google and other evidence that a publication or business has an identity outside an individual optimized page.
Google has not published a policy saying strong direct traffic or brand-query volume automatically causes higher rankings. Those signals should therefore be treated as strategic indicators and practitioner inference, not confirmed ranking factors.
The practical point is still useful. A publisher whose entire discovery model depends on generic informational queries is more exposed to changes in the results page than one that also earns direct visits, repeat readers, branded searches, newsletter traffic and recognition outside Google.
The recovery question also needs more nuance than the industry’s update-by-update language sometimes allows. Google’s core update guidance says improvements can sometimes appear within days but may take several months for its systems to reassess fully.
Google also says sites do not necessarily have to wait for the next major core update because smaller core updates happen continuously and are generally not announced. That makes a fixed recovery timetable difficult to defend.
The same is true of update cadence. Google describes major core updates as something it makes several times a year, not something scheduled every eight or twelve weeks. The March and May 2026 core updates arrived relatively close together, but that should not be turned into a rule that future core updates will follow the same interval.
The professional consequence is a change in planning. Organic search can still deserve investment without being treated as a channel whose quarterly output can be forecast from ranking positions alone.
The AI question hangs over all of this, and separating what is known from what is inferred matters. Google describes spam updates as improvements to automated systems that detect spam, including SpamBrain, its AI-based spam-prevention system. That is different from saying every spam update is simply a new SpamBrain model or that all ranking decisions now run through one AI system.
There is also no public basis for assigning a percentage to how many ranking decisions are made by systems that Google’s engineers cannot interpret. Claims at that level go beyond what Google has publicly documented.
What the public record does show is that Google combines announced broad updates with smaller ongoing changes, while AI-generated search experiences increasingly sit between the query and the publisher. For publishers, those two developments make traffic harder to read from rankings alone.
The forward-looking read is therefore less about predicting the next winning tactic and more about recognizing the environment in which those tactics operate.
Informational content is likely to face continued pressure when a search engine can satisfy a simple question directly on the results page. Pages have a stronger reason to offer something that a summary cannot easily replace: primary reporting, original data, specialist expertise, distinctive analysis, a recognizable voice or a transactional purpose.
Brand demand is also becoming more valuable as a hedge, even if the exact relationship between brand signals and Google’s ranking systems remains debated. A reader who searches for a publication by name, visits it directly or subscribes to it is less dependent on the position of a single blue link.
And volatility should be treated as an operating assumption rather than a temporary inconvenience. Google’s own documentation makes clear that search results are dynamic, that smaller updates continue between announced core updates and that improvements do not produce guaranteed or immediate ranking changes.
The useful distinction for growth teams is therefore becoming simpler. Ranking and being findable are related, but they are no longer interchangeable.
A page can rank and still lose the click. A site can accumulate impressions and still lose traffic. A publisher can recover visibility without recovering the business result that visibility once produced automatically.
That is not the end of SEO. It is a change in what SEO can reasonably promise.