Does an AI citation bring clicks to your website?

What happens to traffic when an answer engine cites a page?

For creators and publishers evaluating answer engines, discovery is often expected to mean traffic. The assumption is intuitive: if an engine cites your work, readers will click through to your website.

The little evidence we found points the other way.

In our reading, the reason is the format. A list of search results asks the reader to click; an answer engine gives the answer in prose and lists its sources beside it, so a reader can stop there.

In one report published by ZipTie, one site owner reported 797,444 impressions in Google’s AI Overviews and 7 clicks.

That is one site’s report, and it does not say why so few people clicked.

We have not measured sitewide clicks or revenue across the web, and we cannot promise that publishing text will bring traffic. Nothing we found shows impressions in an AI answer turning into many visits.

One report also leaves a lot open: which questions the answers covered, how much of the page the answer used, and whether some readers came to the site later by another route. We have not measured any of that ourselves.

Who captures the visitor when an engine references a video?

For video creators, the referral path is further complicated by where clicks land.

When answer engines cite video, they point to the hosting platform. In an analysis by Ahrefs of domains cited in AI Overviews, YouTube led all domains by capturing 22.9% of citations.

Our reading is that when an answer cites your video, the click, if any occurs, goes to YouTube rather than your site.

The creator who made the video gets no visit to their own domain. Relying on platform citations leaves discovery on someone else’s land.

As we noted when discussing why we were wrong about YouTube and AI, answer engines cite video often. But having your video cited on YouTube does not build direct web reach. Any traffic it sends goes to YouTube, not to your own website.

In addition, as shown in our breakdown of which AI engines cite video, different engines treat formats differently. In the one controlled test we found, several assistants cited only the written version. A creator who publishes only video gives those engines nothing to cite.

What are independent creators reporting from search results?

Forum posts point the same way, though they are anecdotes.

On Hacker News, several independent site owners said AI answers in search took much of the traffic their sites used to get. In their account, questions that used to send searchers to their sites are now answered on the results page.

Creators with both formats report the same trend. On Reddit, a creator who writes up their videos as reviews on their own site said the AI answer, built from that site, now shows first, so fewer people visit; they keep doing it anyway.

Similarly, a niche site owner on Reddit said AI Overviews take clicks that used to reach smaller sites.

These are individual forum posts from a few creators, not a statistical study. We have not measured aggregate clicks or commercial revenue from citations.

Why pursue AI citations if direct visits are scarce?

If citations bring few visitors, why should a creator turn video into a written website?

In our reading, the case for being cited is not referral traffic. We cannot promise traffic from a text archive. Instead, the real value is attribution and authority: being named as the source of an answer on engines that cite text.

When an engine answers a question, citing your site names your work as the origin of the answer. In our reading, that credit is the main return of a citation, even without a click.

Furthermore, if a page is to earn a click from an answer, it must provide what the summary cannot:

• The full argument. Detailed analysis, case studies, and depth that cannot fit into an overview snippet.
• Dedicated resources. Tools, templates, products, or consulting that require direct engagement.
• Direct relationships. A newsletter or membership that invites readers to connect directly.

A short summary delivers a brief conclusion, not the depth of a full resource.

We do not claim citations raise brand searches, sales, or subscribers. The benefit is attribution: ensuring that when an engine answers in your field, it names your site as the source.

How should creators approach text archives and citations?

For creators, the practical step is not to abandon written publishing, but to set realistic goals.

Publishing only video gives engines that cite only text nothing of yours to cite, and when engines cite your video, any click goes to YouTube. But expecting large traffic from citations alone is not supported by the evidence we found.

The balanced approach is treating a written archive as an anchor for authority. In The Dark Archive, our study of 1,284 long-form YouTube channels, we analyzed where creator knowledge lives off YouTube and how machines read it.

We also ran our own citation test, and its result is in The Dark Archive report.

Republishing video insights as text on an owned domain makes your ideas readable across both text assistants and search engines. Citations may not bring many clicks, but they can name you as the source.

To see whether your channel has a written footprint that answer engines can read, run the free Dark Matter Check.