X’s Four-Year Climb to the Top 10

November 11, 2026 · 6 min read · By API User

X ended October as the sixth most visited website in the world, with about 4.37 billion visits. That is roughly 200 million more than September, a gain of 4.78%. The number arrived with little ceremony, which is odd given the last four years of predictions that the platform was finished.

Musk completed the $44 billion purchase of Twitter on October 27, 2022. The first year was brutal: advertisers pulled back, staff was cut, verification changed, the API was paywalled, and the company rebranded as X. Similarweb estimated that global web traffic to twitter.com fell 14% year over year in September 2023, US traffic fell 19%, and combined US iOS and Android monthly active users fell 17.8%. The obituary was written in public.

Four years later, X is still a top-10 web property. The October estimate puts it ahead of Reddit, TikTok, and WhatsApp, and behind Google, YouTube, Facebook, Instagram, and ChatGPT. That ranking is a recovery, but it is not a coronation. The race to the top has become a race to stay in the attention business while AI chat and video take more of it.

Visits, users, and estimates

The first thing to untangle is the unit. A visit is a session. One logged-in user who opens X on a laptop, clicks a thread, and returns three times can produce several visits. A news article that embeds a post can send logged-out readers to x.com. A bot can generate visits too. The visit count measures sessions; the user count is smaller and harder to estimate.

Similarweb’s ranking is built from estimated web traffic, not from X’s server logs. The company combines panel data, clickstream, and modeling to estimate visits, unique visitors, pages per visit, and time on site. Semrush does something similar with its own panel and models. Both are estimates. Neither is a census.

The vendors disagree in September. Similarweb ranked x.com sixth worldwide, with an average visit duration of 13 minutes 25 seconds and 14.9 pages per visit. Semrush counted 4.17 billion visits and ranked x.com eighth. The October estimate of 4.37 billion visits is consistent with the same trajectory, but the exact rank depends on the vendor, the month, and how domains are grouped.

There is also a platform gap. Web visits do not include time spent inside the X app. A social platform with a large mobile app can look smaller in a web ranking than it is in daily life. The reverse is also true: logged-out browsing, search traffic, and embeds can make web visits look larger than the active user base.

The first year was a stress test

The post-purchase collapse was real, and it was broad. Similarweb’s first-anniversary report found declines in web traffic, ad-portal traffic, and mobile app usage. Advertisers were nervous about moderation. Publishers watched referral traffic shrink. Users tested alternatives. The rebrand to X in 2023 made the break with the Twitter brand explicit, and the API changes cut off a long tail of third-party clients and research tools.

What did not happen was a mass exit to a single rival. Competitors took slices of the audience, especially around election cycles and moderation fights, but none replaced the network. X retained a core of journalists, politicians, traders, sports fans, and developers who need real-time updates and direct access to sources. That core is valuable even when it is noisy.

Musk also became the platform’s most followed account and one of its most reliable traffic drivers. His posts, fights, product launches, and political interventions created the kind of attention that no marketing budget can buy. That is a double-edged asset: it keeps X in the news, and it keeps advertisers and regulators uneasy.

The 2026 rebound is real, but narrow

The monthly series shows the same direction. Semrush counted 4.25 billion visits in July, 4.06 billion in August, and 4.17 billion in September. The October estimate adds another 200 million visits and keeps the platform in sixth place. The trend is a recovery from the summer trough, with one strong month on top.

The ranking is also a reminder of how much the attention market has changed. ChatGPT now sits fifth, above X. Google, YouTube, Facebook, and Instagram remain ahead. X is competing for the same minutes as AI assistants, short video, and messaging apps, not just other social networks. A top-10 ranking in that field is a strong position. It is not dominance.

And a traffic ranking is not a revenue ranking. Advertiser demand, subscription revenue, data licensing, and payments are separate businesses. X can be heavily visited and still be less profitable than a smaller platform with better ad targeting. The visit count measures attention, not commercial health.

How to measure popularity without fooling yourself

There is no single number for popularity. Each metric answers a different question:

  • Visits: sessions on the web domain. Good for reach and habit, but easy to inflate with logged-out traffic and multiple visits per person.
  • Unique visitors: estimated individuals. Closer to audience size, but still modeled, and it does not capture app-only users.
  • Monthly active users: a platform-reported or panel-estimated count of people who show up in a month. Definitions vary by company.
  • Daily active users: a better measure of habit, but it says little about how long people stay or what they do.
  • Time spent and pages per visit: measures engagement. X scores well here because threads and video encourage long sessions, but a long session can also mean confusion or doomscrolling.
  • App downloads and app DAU: the mobile picture. Sensor Tower, Appfigures, and Similarweb’s app panels estimate this, but app-store rankings are not traffic.
  • Ad reach and revenue: the commercial picture. A platform can have fewer visits and more revenue per user.

The cleanest approach is to use at least two independent sources and to state the unit. Similarweb and Semrush are useful for web comparisons. Comscore and Nielsen panel data are useful for media reach. App analytics firms cover mobile. X’s own metrics are useful for direction, but they are not audited like a public company’s financials, and they do not map cleanly to web visits.

What the next three to five years look like

X’s advantage is speed and access. Breaking news, elections, sports, and market moves still travel through X faster than through most competitors. Grok gives the platform an AI layer that can summarize and search the firehose. Video and creator payouts give it more reasons to stay. Payments could turn attention into transactions if X can clear the regulatory and trust hurdles.

The risks are equally clear. Advertisers can leave again. Regulators can tighten moderation rules. A rival can win a generation of users with a better product. Musk can alienate a large chunk of the audience with a single post. And the AI assistants now competing for attention are not standing still.

The four-year record suggests X is harder to kill than its critics expected. It also suggests the platform is not on a straight line to the top. The 2022 collapse, the 2026 rebound, and the October estimate are all part of the same story: a network with a durable core, a volatile owner, and a business that has to prove it can convert attention into trust.

For a comparison with one of the alternatives that gained ground after Musk’s takeover, see Bluesky’s leadership change and company update.

Bottom line

X is sixth in the world by estimated web visits, with 4.37 billion visits in October and about 200 million more than September. That is a genuine recovery from the first year after the acquisition. The number shows that X remains a top-10 attention property. It does not show that X won the internet. The fight over how attention is measured is now part of the product itself.

Sources and References