NavBoost decrypted: How user clicks influence Google rankings

NavBoost is Google’s ranking algorithm revealed during the antitrust trial and again in the spotlight with the Google Leak. But what is it used for, exactly? How does it use user clicks? We explain everything about how it works.

What you must remember :

  • NavBoost is an algorithm that uses user clicks to improve the relevance of search results;
  • Google has long denied the importance of clicks in rankings, but the antitrust trial and Google Leak have proven their decisive role.

What is NavBoost?

NavBoost is an advanced Google algorithm designed to improve search results based on user interactions with search results pages. Unlike other ranking signals that focus primarily on technical or content aspects of websites, NavBoost emphasizes user clicks. In other words, it takes into account how users interact with search results to determine the relevance and quality of the pages offered.

NavBoost works in remembering user clicks over a 13-month period (18 months before 2017). This memorization allows Google to analyze users’ browsing behavior over an extended period, and thus better understand which pages best respond to user queries.

The main goal of NavBoost is to improve the user experience by providing more relevant and accurate search results, especially for navigation querieswhere the user’s intent is to find a specific site or page.

How does NavBoost work?

The operation of NavBoost is based on analysis of user clicks and other behavioral signals. When users perform searches and click on results, these interactions are recorded and analyzed by NavBoost. Here are some key steps in the process:

  1. Click collection : Every time a user clicks on a search result, this action is recorded. NavBoost retains this data for all requests received over the last 13 months.
  2. Behavior analysis : NavBoost examines click-through rates (CTR), browsing behaviors, site quality, and contextual relevance. For example, if a page receives many clicks for a certain query and users spend time on that page without quickly returning to the SERPs (pogosticking), this indicates that the page is relevant and quality.
  3. Data Segmentation : NavBoost creates separate datasets for mobile and desktop searches, as well as for different geographic locations. This allows search results to be adjusted based on the user’s context.
  4. Integration with other algorithms : Although NavBoost is an important signal, it works in conjunction with other Google ranking algorithms and signals. For example, systems like “Glue” analyze other user interactions on the page, such as scrolling and hovering.

The antitrust lawsuit and the discovery of NavBoost

For many years, Google denied that user clicks impact search rankings. This position was called into question during the 2023 antitrust trial, where important revelations were made regarding NavBoost and therefore the consideration of clicks as an important ranking criterion.

This lawsuit aimed to examine Google’s practices regarding monopoly and unfair competition. During the hearings, several internal documents and testimony revealed crucial information about the inner workings of Google and its ranking algorithms.

History and context

NavBoost has been around since 2005 and has been updated several times. The algorithm is apparently based on a 2004 patent entitled “Systems and methods for correlating document topicality and popularity”.

This patent, co-authored by Amit Singhal, describes how Google uses user interactions to determine the popularity and relevance of documents. Although the patent does not explicitly mention clicksit refers to “user navigation patterns”, which necessarily include clicks.

Testimony from the antitrust trial

Pandu Nayak, vice president of search at Google, testified that NavBoost is one of the most important ranking signals in the 2023 antitrust trial. He explained that NavBoost analyzes user clicks to adjust search relevance scores. pages. According to Nayak, this system helps narrow down the list of documents to a more manageable and relevant selection for users. This testimony contradicts Google’s previous statements denying the importance of clicks.

Impact on ranking

NavBoost changes document scores based on user clicks. For example, a page that receives many clicks for a certain query will see its relevance score increase, which can improve its ranking in search results. Conversely, a page with few clicks or high bounce rates may see its score decrease.

The Google Leak and its revelations about NavBoost

In late May 2024, a massive leak of 2,500 internal Google documents revealed important information about NavBoost, shedding light on its central role in ranking search results through user clicks. Here are the key points:

  • Collection of click data : NavBoost collects user click data from the Chrome browser. These clicks are used to evaluate the relevance and quality of the results pages.
  • Analysis of user behavior : The documents show that NavBoost analyzes not only the number of clicks, but also their duration. Long clicks (where the user stays on the page for a long time) are weighted positively, while short clicks (where the user quickly returns to the SERPs) are considered negative.
  • Analysis of “overwritten” clicks : By overwritten clicks, Google means clicks that it judges to be less reliable or relevant. Uncrushed clicks are those that are considered high quality. NavBoost also takes this into account.
  • Click-based re-ranking : NavBoost uses click data to readjust page rankings in the SERPs. A page receiving a lot of long or non-clicks will see its relevance score increase, thus improving its ranking.
  • Data Segmentation : NavBoost segments click data by geography and device type (mobile vs desktop) to further refine the relevance of search results.

What this changes for SEOs

The official recognition of clicks as an important ranking factor has implications for SEO professionals, although most already thought that clicks were taken into account by the search engine, despite his statements. Some general recommendations for optimizing a site based on the NavBoost algorithm:

  • create a logical and well-organized site tree, so users can easily find the information they are looking for, with clear menus and relevant internal links;
  • work on your UX: design a site with practical navigation and attractive pages to keep users on your site for as long as possible;
  • optimize the loading speed of your site to avoid losing your visitors before they can even access your page;
  • ensure that your site is responsive and optimized for mobile display, to avoid losing visitors who find your interface impractical on a smartphone;
  • respond well to the user search intent to avoid short clicks and therefore a high bounce rate which would send the wrong signal to NavBoost.
-

-

PREV The RN wants to ban “sensitive jobs” for dual nationals
NEXT Vivo Ernergy will market Loukkos gas produced by Chariot