Kyle Eggleston’s AI Overview and AI Content Studies Explained
What Eggleston’s Two Studies Measured
Chicago-based SEO consultant Kyle Eggleston has published two original research reports examining how Google’s AI systems treat written content. Both are first-party studies, meaning the data comes from his own collection rather than a third-party tool’s published index. The first, titled How to Rank in AI Overviews, looks at which pages Google’s AI Overviews cite. The second is a case study of a website built and written entirely with AI tools, tracked through several Google algorithm changes.
Several terms matter for reading the findings. An AI Overview is the generative summary block Google places above traditional results for some queries, and it typically includes outside links. The organic top ten refers to the first page of standard blue-link results. A SERP snapshot is a captured record of what the results page looked like at a specific time, which is why the date range of the capture matters. Domain Rating, a metric from Ahrefs, estimates the strength of a site’s backlink profile on a scale from zero to one hundred. Core updates and spam updates are broad changes to Google’s ranking systems, with spam updates aimed at manipulative or low-value content.
The first study covered 400 US queries drawn from a pool of 3,804 keywords across ten verticals. After removing Shopping cards and dead redirects, it analyzed 1,561 outside citations. The second study followed one new domain from April 27 to September 13, 2026, and reported 1,978 clicks and 127,262 impressions in Google Search Console. Understanding these basics makes it easier to judge how much weight each result can carry.
Key Findings, Benefits, and Caveats
The AI Overviews data shows a strong relationship between organic position and citation. About 51 percent of citations pointed to URLs already in the organic top ten, and that rose to roughly 64 percent when counted by domain. The page holding the number one organic spot appeared in 56 percent of the overviews studied, while the page in eighth place appeared in fewer than 10 percent. Nearly every cited page from outside the top ten still ranked somewhere in the top 100 for at least one keyword. For practitioners, this suggests that AI Overview visibility is not a separate game with unrelated rules.
The platform breakdown also stands out. YouTube led all domains with 265 citations across 145 overviews, which supports the case for video as a citable format. Reddit accounted for 5.1 percent of citations overall but did not appear in health-related results at all, a reminder that source patterns vary by topic.
The AI content case study challenges the belief that AI-written text is automatically penalized. The site weathered the May 2026 core update and the June and August 2026 spam updates without a visible ranking drop, despite having a Domain Rating of zero and no dofollow backlinks. Eggleston argues the more useful question is whether content serves readers, not which tool produced it.
He is careful about limits. The case study is a single anonymized site, and the AI Overview data is a one-day snapshot of US desktop results. Both findings describe correlation and do not prove that updates caused the traffic gains or that AI writing will work everywhere.
Practical Takeaways for Marketers and Site Owners
The most actionable lesson is to treat organic ranking as the foundation. If your target pages already sit in the top ten, they have a materially better chance of appearing in AI Overviews. Audit the queries that matter most, check which URLs are cited, and compare those against your own positions. Where a competitor is cited and you are not, look at its depth of answer, clarity of structure, and freshness before assuming the gap is technical.
For content creation, an AI content generator can speed up drafting, outlines, and research summaries, but the case study points toward human review as the differentiator. Add firsthand experience, verify facts, and make sure each page answers a specific reader question. Visual assets can support this work too. An AI Post Images Generator can produce custom graphics for articles or social posts, though you should check image accuracy, licensing terms, and whether the visual adds real information.
Backlinks deserve a measured approach. Eggleston notes that his site’s performance was not explained by link building, but that does not make links irrelevant for most sites. If you use an auto backlink tool, prioritize relevant, editorially earned placements over high-volume automated links, which can invite quality problems. Finally, track results in Google Search Console, compare impressions and clicks across updates, and document what changed. Your own data will be more useful than any single study’s headline.
