Direct Traffic and Popularity: Correlation, Not Causation
▼ Summary
– Cyrus Shepard’s AI citation ranking factors study sparked debate on the distinction between correlation and causation, a caveat he explicitly made.
– Direct traffic is a symptom of good performance, not a primary ranking factor; treating it as such can lead to misguided tactics like purchasing bot traffic.
– High direct traffic typically indicates a strong brand, correlating with genuine ranking factors like brand searches and quality backlinks.
– Google’s NavBoost and Glue systems use user interaction signals, not raw direct traffic volume, to gauge relevance and authority.
– Popularity is defined as a sign of brand strength, such as autocompletes and bookmarks, and Google may use Chrome data to train AI models rather than as a direct ranking factor.
Last week, Cyrus Shepard released a study on AI citation ranking factors that quickly sparked debate across X, LinkedIn, and several private WhatsApp groups I follow. The conversation centered on a crucial distinction: what actually drives rankings versus what merely correlates with them. This is especially tricky given that most SEO and AI studies involve numerous variables and layers of complexity that are nearly impossible to untangle. Let me be clear , this is not a critique of Shepard’s work. The study is strong, and he explicitly acknowledges the correlation-versus-causation caveat himself.
This discussion reminded me of earlier ranking factor studies that suggested direct traffic is a significant traditional SEO ranking signal. Back then, those studies drew heavy criticism. That criticism resurfaced recently when Google’s DOJ trial documentation revealed a “popularity” signal, reigniting the debate online.
It makes logical sense that direct traffic could be part of how Google measures popularity through Chrome. The company uses Chrome data to discover new websites. It also evaluates a page’s quality based on how users behave after clicking through. But the precise mechanics , the atomic-level weighting of these variables , remain proprietary and undisclosed.
The Correlation Between Direct Traffic and Popularity
Most SEO professionals agree that direct traffic is a symptom of strong performance, not a primary ranking driver. Treating it as a direct ranking factor creates a dangerous feedback loop. It encourages shallow, low-effort tactics like purchasing bot traffic in a misguided attempt to inflate popularity. After all, it is entirely possible to have high direct traffic volumes alongside poor SEO results.
A broader perspective shows that high direct traffic typically signals a strong brand. It correlates with genuine ranking factors such as numerous branded searches, high-quality backlinks, and robust social engagement. These are the true causes of strong rankings. The direct traffic is simply a measurable reflection of the brand’s overall health and success , a rising tide lifting all ships.
If Chrome data were a direct ranking factor, a sudden spike in browser activity on a specific URL would immediately boost it in search results. That would be an easily exploitable loophole. Google actively works to eliminate obvious manipulation of search rankings, and such an exploit would have been shut down years ago.
Other Takeaways from the DOJ Files
The DOJ documents also shed light on NavBoost and Glue, two specialized systems within Google’s infrastructure. These systems focus on user interaction signals rather than raw direct traffic volume.
NavBoost analyzes historical clickstream data and user behavior on search results pages to determine which pages are most relevant for specific queries. It essentially acts as a memory bank of what users have found helpful over time. While NavBoost focuses on organic results, Glue applies the same user interaction principles to all other SERP features , knowledge panels, video carousels, image packs, and featured snippets. Together, they allow Google to assess a site’s authority based on how users engage with it within the search ecosystem, regardless of where that traffic originated.
So, What Is Popularity?
Based on what we know from official and unofficial sources, research, and the collective wisdom of the SEO community, popularity can be defined as a sign of brand strength. It manifests in user behaviors like autocompletes and bookmarks. It correlates with high rankings because it naturally aligns with the various signals that make a page rank well.
Google may avoid using Chrome data directly as a ranking factor. Instead, it might use that data as a training set or validation tool for its AI models. We simply do not know, and it is unlikely we will ever prove or disprove this through research.
Thank you to Ryan Jones, Mark Williams-Cook, Chris Green, Gerry White, Kristine Schachinger, Charlie Whitworth, Emina Demiri Watson, and anyone else I missed for the engaging weekend discussions on this topic.
(Source: Search Engine Journal)
