Jia Lissa
The SEO community has long been obsessed with the "Holy Grail" of ranking factors—the elusive signals that dictate why one site ascends to the top of the SERPs while another languishes in obscurity. Last week, the release of Cyrus Shepard’s comprehensive study on AI citation ranking factors sent shockwaves through the industry. The discourse, which played out across X, LinkedIn, and private professional circles, served as a potent reminder of the SEO industry’s persistent struggle to distinguish between genuine causation and mere correlation.
While Shepard’s study is a masterclass in data analysis, it also inadvertently reignited a perennial debate: Does direct traffic act as a ranking factor, or is it merely a byproduct of a healthy, authoritative brand? As we navigate an era where AI-driven search and traditional organic algorithms collide, understanding the distinction between these signals is more critical than ever.
The Chronology of a Controversy: From DOJ Files to Modern SEO
The debate over whether direct traffic influences search rankings is not new; it is a recurring theme in the history of search engine optimization. Years before the current obsession with AI citations, industry studies frequently pointed toward a positive correlation between direct traffic volume and high organic search rankings.
These earlier studies were met with significant skepticism, and the discourse was recently invigorated by the revelations emerging from Google’s Department of Justice (DOJ) antitrust trial. The documents released during the trial provided an unprecedented look into the "black box" of Google’s ranking systems, specifically highlighting a "popularity" signal.
The Role of Chrome Data
The revelation that Google utilizes Chrome data to assess website popularity shifted the narrative. We know that Google uses browser activity to discover new URLs and evaluate page quality based on user interaction post-click. However, the internal "atomic levels" of this process—the precise weighting of variables such as dwell time, pogo-sticking, and bounce rates—remain locked behind Google’s proprietary walls.
The industry’s collective leap of logic—assuming that if Google uses Chrome data to track popularity, then direct traffic must be a ranking factor—has fueled a dangerous, persistent narrative. This misconception has led to a misinformation loop, where practitioners, desperate for quick wins, invest in low-effort tactics like purchasing bot traffic to artificially inflate direct visits. The reality, however, is far more nuanced.
Supporting Data: The Illusion of Direct Traffic as a Driver
The fundamental flaw in treating direct traffic as a direct ranking signal is the assumption of causality. In the world of SEO, "all ships rise in high tides."
The "High Tide" Effect
High direct traffic is rarely the cause of high rankings; it is a symptom of brand dominance. A company with a massive, loyal customer base will naturally have high direct traffic, significant brand-name searches, a robust backlink profile, and high social engagement. These are the elements that Google’s algorithms are explicitly designed to reward.
If direct traffic were a primary ranking signal, the system would be laughably easy to exploit. A sudden, artificial spike in browser activity on a specific, low-quality URL would theoretically propel it to the top of the SERPs. If this were true, black-hat SEOs would have weaponized browser automation to dominate search rankings years ago. The fact that Google has consistently stamped out such manipulations suggests that their systems are far more sophisticated than simple traffic-counting.
The Evidence of Disconnect
Recent analysis indicates that there is often zero correlation between sharp, short-term spikes in direct traffic and organic search performance. When a site experiences a "viral" moment that drives direct traffic but lacks the underlying technical and content-based authority, organic rankings remain stagnant. This confirms that Google’s ranking systems are looking for deeper, more sustainable indicators of quality than raw traffic volume.
Official Responses and Internal Systems: NavBoost and Glue
To move beyond the speculation of the SEO "hive mind," we must look at the specific infrastructure Google has disclosed through legal filings. The most significant of these are the internal systems known as NavBoost and Glue.
NavBoost: The Historical Memory
NavBoost acts as a repository of historical clickstream data. It does not simply look at how many people visit a site; it analyzes how users behave within the search results. It effectively functions as a collective "memory" of which pages provided the most satisfactory answers to specific queries. By aggregating this behavioral data, Google can determine which results are worth promoting without relying on the crude metric of direct site visits.
Glue: Expanding the Ecosystem
While NavBoost focuses on the traditional blue-link organic results, the "Glue" system extends these user-interaction principles across the entire search experience. Glue governs how Google evaluates the authority of a site within knowledge panels, video carousels, image packs, and featured snippets.
These systems are designed to gauge authority independent of the user’s traffic source. Whether a user arrives via a direct link or a search result, Google’s priority is assessing the value of that interaction. This confirms that "popularity" in Google’s eyes is not about where the traffic comes from, but about the quality of the engagement that follows.
Defining Popularity in the Age of AI
If direct traffic is not the metric, what constitutes "popularity" to an algorithm? Based on current research and the insights of leading SEO experts, popularity is best defined as a constellation of brand-strength indicators.
The Pillars of Modern Popularity:
- Brand Searches: High-intent queries where users search for a brand name directly.
- User Behaviors: Autocomplete associations (where a brand is suggested alongside a generic keyword) and browser-level bookmarking activity.
- Entity Association: How often a brand is mentioned in context with specific topics, which AI models use to validate expertise and topical authority.
Google likely does not use Chrome data to count "hits" on a page to determine rank. Instead, it uses this massive dataset as a training ground to validate its AI models. Chrome data provides the "ground truth" that helps Google’s machine learning systems understand what users actually find helpful, allowing the algorithm to refine its organic ranking signals over time.
Implications for Future SEO Strategy
The transition from thinking about "ranking factors" to "brand health" is the most important shift a modern SEO practitioner can make.
Moving Beyond Superficial Tactics
The "misinformation loop" regarding direct traffic is a distraction from the work that actually yields long-term results. Purchasing traffic or engaging in click-manipulation schemes is not only ineffective but risky. Google’s sophisticated AI-driven systems are increasingly capable of identifying patterns of manipulation.
The Holistic Approach
If you want to improve your rankings, focus on the variables that drive legitimate popularity:
- Invest in Brand Building: Make your brand a household name within your niche so that users seek you out directly.
- Prioritize User Experience: Use the lessons from NavBoost to your advantage. Ensure your content is so useful that it becomes the "obvious" choice for users, resulting in high dwell time and repeat visits.
- Strengthen Your Entity: Focus on E-E-A-T (Experience, Expertise, Authoritativeness, and Trustworthiness). Ensure your site is linked, cited, and recognized as an authority by other reputable sources in your industry.
Final Thoughts
The study published by Cyrus Shepard regarding AI citations serves as a vital touchstone for the industry, but it must be read with the understanding that SEO is not a science of isolated variables. It is a complex, multifarious ecosystem.
Direct traffic is a vanity metric that happens to correlate with success, but it is not the engine of that success. As we move further into the age of AI-integrated search, the winners will not be those who try to game the traffic stats; they will be the ones who build brands so robust that their popularity becomes an undeniable, objective truth that the search algorithms have no choice but to recognize.
Special thanks to the community members and experts, including Ryan Jones, Mark Williams-Cook, Chris Green, Gerry White, Kristine Schachinger, Charlie Whitworth, and Emina Demiri Watson, whose rigorous debate and critical thinking continue to push the SEO field toward a more sophisticated understanding of search dynamics.
Rifan Muazin
Ammar Sabilarrohman
