AEO needs new metrics

Search marketers are starting to treat answer engine optimization as a measurement problem, not just a content problem. New guidance says classic search metrics like clicks and rankings miss how artificial intelligence systems pick and cite sources. (hubspot.com) HubSpot said on April 22 that teams should track metrics built for AI answers, including visibility in generated responses, citation frequency, share of voice, and sentiment. Its new guide separates those from standard search engine optimization indicators such as organic traffic and click-through rate. (hubspot.com) Search Engine Land has been making the same case in 2026, framing generative engine optimization around whether ChatGPT, Google AI Overviews, Perplexity, Claude, and Copilot retrieve and cite a brand at all. Its recent guides describe the core job as becoming “the source AI engines cite” when they assemble an answer. (searchengineland.com) The shift follows changes in the products people use to search. Google said its AI Overviews and AI Mode are designed as a “fluid experience” with follow-up questions and prominent links, moving some discovery from a list of blue links into generated summaries. (blog.google) Google’s own documentation still tells publishers that standard search engine optimization practices matter for AI features, but it now has a separate guide for how sites appear in AI Overviews and AI Mode. That split has given marketers room to argue that performance reporting needs its own layer too. (developers.google.com) The new metrics are trying to capture a different funnel. Instead of asking whether a page ranked No. 3 and won a click, marketers are asking whether a model selected the page, cited it in the answer, and surfaced fresh information signaled by update dates and page maintenance. (hubspot.com) Software vendors are already packaging those ideas into dashboards. HubSpot says its AEO product tracks a visibility score, prompt tracking, citation analysis, and competitive share of voice across ChatGPT, Gemini, and Perplexity. (hubspot.com) Microsoft has also started exposing AI-specific reporting. Bing Webmaster Tools added an AI Performance public preview in February 2026 that shows when a site is cited in Microsoft Copilot, Bing AI summaries, and some partner integrations, along with the URLs referenced over time. (blogs.bing.com) That leaves marketers with two scoreboards running at once: one for traffic from conventional search results, and one for visibility inside generated answers. As answer engines keep inserting themselves between the question and the click, the second scoreboard is getting harder to ignore. (developers.google.com)