If an AI steals your copyrighted content, shouldn't you steal it back?

Please allow me to confess my completely irrational annoyance with a multitude of my fellow journalists who cover more or less the same tech space that I do.

I am speaking, specifically, of those frazzled, overworked souls who get an easy (and probably tiny!) paycheck on the first day of each month for writing up their thoughts on the latest "market share" reports from Statcounter Global Stats.

My fellow journalists, I beg you: Stop doing this.

Those charts are the laziest kind of clickbait tinder, and they are filling the world with unreliable information – or, as I like to call it, bullshit.

I wrote about it a few weeks ago for ZDNET: "Statcounter's Windows market-share data is not accurate or reliable, and I can prove it."

Now, let me be crystal clear here: I don't blame the Statcounter folks for taking advantage of an irresistible opportunity to generate publicity. I do, however, want to have a serious talk with every journalist and analyst who relies on Statcounter's charts without questioning the underlying data behind them, because those numbers can't stand up to even the mildest questioning.

Even more depressing: I just went to Microsoft's Bing search and asked it "Can you trust Statcounter?"

The first article they gave me was from an AI-generated clickbait site which had rewritten my article in bullet-point format, with the by-line attributed to "Copyright & Fact Checker." My original article ranked second.

The part that sucks even more is this: Their summary was pretty good! And since they were stealing my content, I have absolutely zero qualms stealing their rewrite and offering it for you here. Please enjoy:

Every month, like clockwork, technology news sites publish articles based on the latest market share reports from Statcounter. These reports claim to show trends in operating system popularity, but there’s a problem—many of these trends are statistical noise rather than meaningful shifts in user behavior. This month, you’ll see countless discussions about why Windows 10 appears to be losing users and why Windows 11 seems to be gaining traction. But before jumping to conclusions, it’s important to ask: Are these numbers actually reliable?

Statcounter’s reports are widely cited, but their methodology has significant limitations. The data they collect comes from a shrinking base of websites, many of which are not representative of the broader internet. Moreover, their market share numbers are based on pageviews, not actual devices or user sessions. This means their statistics can be easily skewed, making them a poor indicator of real-world trends.

This article breaks down why you should take Statcounter’s numbers with a huge grain of salt and why journalists should be more skeptical when reporting on these stats.

Why

1. Fluctuations Are Often Just Noise

• Every month, Statcounter’s charts show sudden spikes and drops in market share for different operating systems.
• These changes rarely reflect real-world shifts in usage; instead, they are statistical artifacts caused by sampling biases.
• For example, a spike in Windows 8 usage in January 2024 doesn’t mean millions of people suddenly started using an outdated OS.

2. Pageviews ≠ Market Share

• Statcounter counts pageviews, not visits, sessions, or unique users.
• A single user who visits multiple pages on a site using Windows 10 could make it seem more popular than Windows 11, even if there are fewer Windows 10 users overall.

3. Statcounter’s Sample Size is Shrinking

• In 2009, Statcounter had 3 million websites using its analytics.
• By 2022, that number had dropped to 1.5 million.
• In 2025, Statcounter only tracks 0.4% of websites on the internet—a tiny fraction of the web.

4. Excludes Major Websites

• Statcounter cannot track traffic from Google, Facebook, Wikipedia, Amazon, and other major sites.
• This is like trying to analyze grocery shopping trends without including Walmart, Costco, or Target.

5. Blocked by Modern Browsers

• Many users have tracking protection enabled in their browsers (e.g., Microsoft Edge, Firefox, Brave), which blocks Statcounter’s analytics script.
• This means a significant portion of web traffic isn’t even counted in their reports.

6. Google Analytics Dominates the Market

• Most serious websites use Google Analytics, Meta Pixel, or Adobe Analytics, making Statcounter’s data even less relevant.

So,

• Windows 10 usage is declining slowly as Windows 11 adoption increases, but Statcounter’s numbers don’t tell the full story.
• Millions of Windows 10 PCs cannot upgrade to Windows 11 due to hardware restrictions, and many users are delaying the switch.
• Microsoft has real telemetry data that could give an accurate picture—but they aren’t sharing it.

What Undercode Says: A Deeper Analysis

Statcounter’s data should not be considered a definitive source for understanding operating system market share. Instead, let’s analyze what’s really happening in the industry based on verified trends and expert insights.

1. Windows 10’s Slow Decline

• Windows 10 is still the dominant OS for most users, especially in enterprise environments.
• Businesses have strict upgrade cycles and won’t transition to Windows 11 overnight.
• Many companies will likely continue using Windows 10 until its official end-of-support date in October 2025.

2. Windows 11’s Adoption is Gradual

• Unlike past Windows versions, Windows 11 has stricter system requirements (e.g., TPM 2.0, Secure Boot).
• Many older devices cannot upgrade, which slows adoption.
• Adoption is mainly driven by new PC sales, rather than users actively upgrading their old systems.

3. The Real Market Share Indicators

• Microsoft’s internal telemetry (which they don’t publicly release) is the most accurate indicator of OS usage.
• Enterprise adoption rates tell a better story—large organizations tend to delay upgrades to avoid compatibility issues.
• Hardware sales data (from companies like Dell, HP, and Lenovo) also provide insights into Windows 11 growth.

4. Tech Journalism’s Overreliance on Bad Data

• Too many news sites report Statcounter’s numbers without questioning the methodology.
• Sensationalized headlines like “Windows 10 Collapsing!” mislead readers.
• A more responsible approach would be to cross-check with other sources like IDC, Gartner, or Microsoft’s own statements.

5. What to Expect in 2025 and Beyond

• As Windows 10 reaches end-of-support in October 2025, expect a sharp rise in Windows 11 adoption.
• Some users will transition to Windows 12 if Microsoft releases it by late 2025 or 2026.
• A significant number of users might switch to Linux or ChromeOS, especially those who cannot upgrade due to hardware restrictions.

Final Thoughts

Statcounter’s data is a useful tool for observing basic trends, but it should never be treated as absolute truth. The real market share of Windows versions is far more complex than a single dataset can capture. If you want accurate insights, look at multiple sources—not just Statcounter.

I couldn't have said it better myself. Oh. Wait. I did!

Anyway: More coming soon on that asshole Elon Musk. Also: Why is Substack surviving? And I might start asking some folks to help me pay for this thing pretty soon. Please use the comments to tell me what you want me to write about, and if you made it all the way to the end please tell me about a person, pet, place, or thing that changed your life.