I also want to mention that I noticed the importance of clicks when I started a PPC Campaign on a website that was not fully optimized. After just two weeks, the website began ranking on the first page of the SERPs, and even in the top three for long-tail keywords. I was surprised then, so yesterday, when I read that “clicks were part of Google’s ranking algorithm,” I had already experienced it.
The Former Googler’s Revelation
Eric Lehman, a former 17-year employee of Google, recently testified as part of the ongoing U.S. vs. Google antitrust trial. Lehman’s statement that “pretty much everyone knows we’re using clicks in rankings” has garnered significant attention. This admission hints at the significance of user behavior in influencing search engine rankings.
However, it’s essential to clarify that clicks alone are not the sole factor determining rankings. Google’s advanced machine learning systems, including BERT and MUM, are becoming even more crucial in evaluating text and providing relevant search results. This suggests that while user data, including clicks, plays a role, it is not the sole determinant of a website’s search engine performance.
The Evolution of Google’s Algorithm
According to Lehman, Google is gradually moving away from heavy reliance on user data towards using machine learning for text evaluation. This shift has been in the making for some time, as demonstrated by Lehman’s email from 2018, where he stated, “Huge amounts of user feedback can be largely replaced by unsupervised learning of raw text.”
The distinction between “user data” and “training data” is also crucial in understanding Google’s approach. Google’s advantage in using BERT, for example, lies in its ability to leverage training data, not user data. This clarification emphasizes that Google’s focus is on improving the quality of search results rather than solely relying on user-specific interactions.
