DEV Community

Alexander Todosuik
Alexander Todosuik

Posted on

We Analyzed 1 Million Keywords to See How AI Is Actually Changing Click Distribution in 2026

Most discussion of AI's effect on search traffic is anecdotal: a screenshot of one site's traffic graph, a single case study, a claim from one agency's client portfolio. A dataset of 1 million keywords gives a very different picture -- not "AI is killing organic traffic" or "AI is fine, nothing changed," but a much more specific breakdown of where clicks are actually being redistributed and where they are not.

The Headline Number Everyone Quotes Is Misleading on Its Own

"AI Overviews reduce organic clicks" is true in aggregate and useless for planning. The actual redistribution is heavily concentrated in specific query types, not spread evenly across all search traffic.

Purely informational queries with a short, factual answer -- definitions, conversions, single-fact lookups -- show the steepest click decline when an AI Overview is present. Queries requiring comparison, opinion, recent data, or a decision between options show a much smaller decline, and in some categories, close to none.

Where the Data Diverges From the Common Narrative

Commercial investigation queries (comparison shopping, "best X for Y" formats, pricing research) are far more resistant to AI Overview click suppression than purely informational ones. The AI can summarize a definition; it struggles to fully replace a comparison the user wants to verify themselves before spending money.

Branded and navigational queries are barely affected -- unsurprising, but worth stating plainly, since some site owners have applied "AI is eating our traffic" broadly to categories where the actual data shows minimal change.

Local intent queries show a different pattern entirely, tied more to Business Profile and map-pack visibility than to organic listings, meaning the AI Overview conversation mostly does not apply to them the way it does to national informational content.

Full breakdown of the query-category data and what it means for content prioritization in 2026: https://yositeup.com/blog/ai-reshaping-google-search-1m-keywords-2026

What This Means for Content Planning

Treating "AI Overviews are hurting my traffic" as a single problem with a single fix (write more FAQ content, add more schema) misses that the actual impact is category-specific. A site heavy on short factual/definitional content is in a fundamentally different position than a site built around comparisons, original research, or transactional intent.

The practical shift for 2026 content planning: audit which category your existing traffic falls into before deciding whether the AI Overview conversation is even the right problem to be solving for your specific content mix.

Top comments (0)