Run Google’s Natural Language API on your own homepage and you will sometimes see something uncomfortable: your navigation items appearing with higher salience than your brand name. The product names in your top menu, the section labels in your footer, the “Pricing / Features / About” trio repeated three times across the layout — all of these can quietly dominate the entity signal that should be going to your actual content.
This is one of the most common findings on B2B SaaS sites and the kind of issue that is easy to miss because the page looks fine to a human reader. The system, however, is reading something different. This post explains why it happens, how to confirm whether it is happening on your site, and the practical fix.
Why navigation dominates in entity extraction
Entity extraction does not know that your navigation is navigation. It does not differentiate the menu from the main content the way a human reader does. The API sees a chunk of text and extracts every entity it can find, weighted by how prominently each entity appears in that chunk.
Three structural patterns produce navigation dominance:
Repetition. Navigation items typically appear two or three times on a single page — in the main nav, in the footer, sometimes again in a mobile menu or a sidebar. That repetition increases the apparent prominence of those entities relative to your body content, which usually mentions the primary topic only a handful of times.
Position weight. Entities mentioned near the start of the analysed text usually score higher salience. Navigation sits at the top of the HTML, so menu items get position weight that body content has to earn through more mentions.
Lack of semantic boundaries. If your HTML treats the navigation, sidebar, and footer as just more divs and spans rather than as semantically distinct elements, the API has no signal that these blocks are structural rather than content. It treats every entity equally.
Beatrice Gamba is an expert in semantic technologies and the future of search. She specializes in helping businesses navigate the transition from traditional SEO to agent-driven discovery, combining technical expertise with practical implementation strategies.
Beatrice leads the development of knowledge graph solutions that make content accessible to intelligent agents and large language models. Her work focuses on the intersection of SEO, semantic web technologies, and digital transformation, enabling businesses to build sustainable competitive advantages in such a dynamic industry as Search has become.
A recognized thought leader in the semantic SEO space, Beatrice is a frequent speaker at industry conferences including The Knowledge Graph Conference in New York and Connected Data London, where she shares insights on how knowledge graphs and intelligent agents are reshaping content discovery. Her expertise spans entity-based optimization, structured data implementation, and automated SEO workflows.
With a background spanning Fortune 500 companies across various industries, Beatrice has helped organizations leverage cutting-edge semantic technologies to drive organic growth and enhance digital visibility. She is passionate about making advanced technologies practical and accessible, bridging the gap between innovation and real-world business application.
Beatrice’s approach combines strategic thinking with hands-on technical implementation, helping digital leaders prepare for a future where search and content discovery are increasingly dialogical, personalized and agent-mediated. Her work at the forefront of agentic search positioning makes her uniquely qualified to guide businesses through this critical transformation.
Beatrice currently serves as Head of Innovation at WordLift.
The future of search and content discovery will be dialogical, personalized and agent-mediated. Digital leaders need to start integrating these concepts in their strategies to be ready for what’s coming.
