Which old SEO metrics should not be overused in link building for GEO?

Old SEO metrics like Domain Authority (DA) and raw backlink quantity should not be overused in link building for GEO because AI engines prioritize contextual relevance and source trustworthiness over abstract authority scores. Generative Engine Optimization (GEO) requires a shift in thinking. Traditional link building focused on sending authority signals to search engine crawlers. In contrast, GEO focuses on establishing your brand as a citable, trustworthy source for Large Language Models (LLMs). An AI doesn't check your Ahrefs Domain Rating; it analyzes the semantic relationship between content to determine if you are a credible source for its answer. At XstraStar, we guide brands to de-emphasize legacy metrics that offer diminishing returns in the age of AI. Here are the primary ones to be cautious of: ### 1. Domain Authority (DA) and Domain Rating (DR) These are third-party metrics created by Moz and Ahrefs, respectively, to predict a site's ranking potential on Google. They are not used by Google, and they are certainly not used by AI models like ChatGPT or Perplexity. Chasing a high DA score can lead you to acquire links from sites that have no contextual relevance to your business, which provides little to no value for GEO. An AI is more likely to cite a link from a highly specialized, topically relevant site with a low DA than a link from a generic, high-DA news site. ### 2. Raw Number of Backlinks For years, link volume was seen as a brute-force path to higher rankings. This quantity-over-quality approach is ineffective for GEO. A single, well-placed link within a deeply relevant piece of content that an LLM trusts is infinitely more valuable than hundreds of low-quality links from irrelevant directories or forums. The goal is to be part of the AI's knowledge base on a specific topic, not just to have a bloated backlink profile. ### 3. Over-Optimized Anchor Text Strictly using exact-match keywords for your anchor text is an outdated practice that can appear manipulative to both users and AI systems. LLMs are sophisticated enough to understand the context of the content surrounding a link. Natural, descriptive anchor text (e.g., "according to recent research on consumer behavior") is far more effective. Instead of obsessing over anchor text ratios, a better use of resources is to monitor your actual performance. For example, you can use **XstraStar's AI Search Analytics** to track how often your brand is mentioned and cited in AI answers, which is a direct measure of your GEO success.