From People Also Ask to AI Search — Reading the Signals That Matter | Mark Williams-Cook

Mark Williams-Cook (AlsoAsked, Candour agency) and host Gianluca Fiorelli discuss how AI search reshapes SEO practice. Williams-Cook frames the shift as evolutionary not revolutionary: the core activities (technical health, content quality, coverage/authority) remain identical, but their relative weightings change. The conversation's most structurally interesting idea is "intent proximity" — People Also Ask data represents the next-most-likely question a searcher will ask, forming a manual version of what AI search does automatically via query fan-out. Google internally measures Time To Result (how fast a user's intent is fulfilled), which explains why AI search will win long-term: it reduces friction for comparable outcomes. Williams-Cook reveals he's building queryfan.com — an open community database of real AI prompts and their associated query fan-outs, letting SEOs see what the LLM actually searches for when triggered by specific prompts. This is the first tool designed to reverse-engineer the retrieval layer rather than the generation layer. On the decay of traditional signals: if users migrate to AI search (80%+ searches inside LLMs), they stop clicking SERPs and stop linking to websites. The link graph and click data that power current ranking will decay. LLM companies releasing browsers (Arc, etc.) signals their need for interaction data to replace these dying signals. Williams-Cook estimates this is "quite far away" because no one has a working solution for LLM-scale answers without web graph metrics. Fiorelli contributes a key operational insight: Google's SERP features (topic filters, People Also Ask, People Also Search) are not just ranking features — they're a real-time map of how Google categorizes topics. The order of tabs in the search menu (left-to-right in Western markets) reveals Google's perceived dominant intent. This is free competitive intelligence most SEOs ignore. The Google API leak discussion reveals the site quality score as the most important discovery: below a threshold score, your content is invisible regardless of quality. The score is calculated primarily from user click data; when unavailable, Google builds phrase models and compares them to known-quality sites. Amazon.com scored average; university FAQ subdomains scored highest (because 100% of their search traffic was intentional navigational queries). Implication: brand search volume is now a survival metric, not a vanity metric.