Channel sheet · CH-28 · gain 4 min · logged November 13, 2026

Content & SEO in the AI EraDirect input

Top Million Website Visits Remain 12.2% Below 2022 Peak

Similarweb data put Q3 2026 visits to the world’s top million websites 12.2% below the Q1 2022 peak. Rank concentration barely shifted, but app use and regional gaps remain.

By Nathan Brooks4 min read876 words

Signal notes

  1. Q3 2026 visits to the top million websites totaled 1.970 trillion, 12.2% below the Q1 2022 peak.
  2. The Q1 2022 peak was 2.243 trillion; the Q1 2026 low was 1.914 trillion, down 14.7%.
  3. The top 100 sites held about 36% of visits in Q3 2026, versus roughly 33% in Q4 2021.
  4. Visits per internet user fell 29.2%, but the user estimate includes revisions to India and China data.
  5. SparkToro corrected an earlier 24% decline headline after Fishkin said its graph and data were incorrect on Nov. 10.
Visits To The Top Million Websites Fell 12%, Data Shows
Input monitorVisits To The Top Million Websites Fell 12%, Data Shows — AS_Photography / pixabay

By Nathan Brooks

Visits to the world’s top million websites reached 1.970 trillion in Q3 2026, 12.2% below their Q1 2022 peak, according to Similarweb data shared by SparkToro. The latest total was above the Q1 2026 low of 1.914 trillion, when the decline from the peak had reached 14.7%.

What do the visit totals show?

SparkToro co-founder and CEO Rand Fishkin asked Similarweb for worldwide desktop and mobile web-panel visits to the top million sites, plus a breakdown by rank. The chart shows totals staying roughly level through 2023, dipping in early 2024 and falling again in the first half of 2025.

The peak came in Q1 2022 at 2.243 trillion visits. Fishkin said the project began with his theory that the largest sites, largely Big Tech properties, were taking more of the web’s traffic. The measured rank split barely changed, so the results do not establish a large shift toward those companies.

Did the biggest sites take a larger share?

Similarweb grouped the million sites into five ranking bands, from ranks 1-10 and 11-100 down to 100,001 through 1,000,000. The first two bands cover the leading 10 sites and the next 90.

Together, the top 100 accounted for roughly 33% of visits in Q4 2021 and 36% in Q3 2026. Each band below the top 100 ended with a slightly lower share than it held at the start.

Those shares matter less than they first appear because the traffic base shrank. A roughly stable or slightly smaller share translated into fewer visits for sites outside the top 100, not a gain in traffic for lower-ranked sites. Fishkin called the small change in the top-100 share minor: “stagnation is a more accurate description of the trend.”

How reliable is the visits-per-user figure?

SparkToro also reports a 29.2% fall in visits per internet user. It calculates that measure by dividing Similarweb’s total visits by DataReportal’s estimate of quarterly global internet users. The table lists 4.95 billion users in Q4 2021 and 6.15 billion in Q3 2026, an increase of about 24%.

The per-user calculation has a reporting break in its history. The user series jumped from 5.65 billion to 6.04 billion during Q3 2025. DataReportal’s October 2025 report said the jump came largely from significant revisions to official internet-use rates in India and China, and warned against treating it as hundreds of millions of people coming online in 90 days.

SparkToro did not mention that reporting change in its post. Its 24% growth estimate compares data from before and after the change. As a result, the 29.2% per-user decline rests on a series with a known reporting change. The 12.2% decline in total visits does not use those user estimates.

What does the panel leave out?

Similarweb’s panel tracks visits in desktop and mobile browsers, but it does not count app use. Fishkin wrote that consolidation by Big Tech is likely more common in apps. That makes the report’s narrow browser focus a poor test of his broader theory about where traffic is concentrating.

The data are worldwide, and Fishkin said it is plausible that traffic to U.S., U.K. and EU sites declined while regions with growing internet access, such as Asia and Africa, grew. The post provides no regional breakdown and does not state the panel size, so neither pattern is tested.

Fishkin did not assign a cause to the decline. His list includes app-only use, reliance on search-engine results, AI tools, social platforms and behavior changes after the COVID-19 pandemic. Those are hypotheses, not findings supported by a regional or device-level breakdown in this report.

What changed after publication?

SparkToro revised the article after publication. On Nov. 10, Fishkin said on X that the prior version “had an incorrect graph/data.” The earlier headline put the visit decline at 24%, and the URL still displays that obsolete figure, which should not be confused with the separate 24% rise in estimated internet users. The current data and headline use 12.2%.

What can operators use from the report?

SparkToro presents the study as an international benchmark for operators trying to place a multi-year traffic decline in their own reports. The most defensible headline remains the 12.2% drop in total visits: rank-based percentages stayed close to steady while the number of visits fell.

The post contrasts this measure with earlier zero-click data from June, which tracked clicks per U.S. Google search. Its own measure counts browser visits to the top million websites worldwide, regardless of where those visits originated. SparkToro sells audience-research software, and Fishkin co-authored the book “Zero Click Marketing.”

Until a regional breakdown is available, Fishkin says the figures should serve as global context, not a benchmark for any single market. He plans to compare them with other sources and revisit the numbers every year or two.

The defensible conclusion is narrow: browser visits to the million biggest sites contracted, and the report does not show whether those visits moved to apps, other regions or other online destinations.

via searchenginejournal.com (Source)

Filed under

  • web-traffic
  • traffic-measurement
  • similarweb
  • zero-click-search
  • big-tech
Share this article:

More from Nathan Brooks

Nathan Brooks

Show full bio

Correspondent covering regulation, policy and their impact on marketing professionals running AI in the stack at Mart Signal.

154 articles

Bus out

‹ Previous articleNext article ›