Programming Pages Lost a Median 57.9% of Their Wikipedia Readers Since 2022, the Steepest Drop of the 12 Topic Groups We Tracked

Summary

Picture a small software agency that publishes tutorials to attract clients. Its post on how a hash table works used to bring in a few hundred developers a month. Now developers paste the question into a coding assistant inside their editor and never open a browser. Wikipedia's programming pages give a sense of how far that shift may have gone.

We tracked 331 Wikipedia articles on programming languages, algorithms, data structures and core coding concepts. The median article lost 57.9% of its readers between 2022 and the latest 12 months. Two out of three lost more than half.

Of the 12 topic groups we measured across health, law, finance, home services, cars, travel and more, programming fell the furthest and fell first. Developers were among the earliest heavy users of AI assistants, and the reading data fits that pattern.

What we measured

We built a seed list of programming topics and matched each to its English Wikipedia title. For each article we compared desktop plus mobile web views in calendar year 2022 with October 2025 to September 2026.

We excluded articles with fewer than 1,000 views in 2022 or a change of more than 20 times either way, which usually signals a rename. We report medians because a few very large pages had unusually high 2022 traffic. Title-level data is monthly, so a weekday vs weekend split per article was not possible; we used desktop vs mobile instead.

Programming fell further than any other topic we tracked

Median change per article across our 12 seed-list studies:

The groups at the bottom of the chart are ones people still look up by name, such as a car model or a company. The groups at the top are ones people look up to understand how something works.

Concepts and data structures fell more than languages

Within programming, the subgroups split this way:

Concept pages like While loop (down 88.5%), Front-end web development (down 85.3%) and Linear search (down 83.8%) are textbook questions that a coding assistant answers in one line. Language pages held up somewhat better, and a few newer or niche languages grew, including Zig (up 135.3%) and Haskell (up 53.0%).

Desktop readers left first and fastest

Programming is a desktop topic. In 2022, 77.8% of views of these pages came from desktop.

Desktop views fell 62.7% and mobile web views 49.6%. Desktop's share of views dropped from 77.8% to 72.1%. Unlike most topics, which held steady in 2023, programming lost a fifth of its readers that first year.

What this means

If you sell to developers, whether that is a SaaS tool, an API or a dev agency, generic tutorials are no longer a reliable way to be found. The questions those posts answered are being handled inside coding assistants and chat tools. For a developer tool, the pages an AI SEO agency should care about are docs, changelogs and integration guides, not another tutorial.

What still gets cited is specific: your product's docs, real benchmarks, migration guides and comparisons that only you can write. Make those easy for AI tools to read and quote. Run a free AI visibility audit to see whether assistants mention your product when developers ask for options; that test is where an AI search optimization agency would start too.

Limits

Wikimedia pageview counts include some automated traffic that its filters do not catch, and Wikimedia changed how it detects bots in 2025, which lowers counted views from that point on. Some of the change in any period may reflect better bot filtering rather than fewer human readers.

We count views of titles that had at least 5 views in a given month on a given site, which drops a very small tail of rarely read pages.

Some large language pages show unusually high 2022 traffic that may include automated requests, which is why we lead with medians.

The seed list was built by hand and leans toward topics taught in computer science courses and common in job listings.

Comparison groups come from our other seed-list studies and use the same method.

Frequently asked questions

Related studies

• Wikipedia readership in the AI answer era'What Is' Wikipedia Pages Lost a Median 41.9% of Readers
• Wikipedia readership in the AI answer eraWikipedia Traffic Since ChatGPT: English Views Down 8.1%
• Wikipedia readership in the AI answer eraSpanish Wikipedia Views Fell 36.3%; English Fell 8.1%
• Wikipedia readership in the AI answer eraPopular Wikipedia Pages Lost 24.5%; Niche Pages Gained 12.2%