Demo · Meaning-Flow · SEO, AEO & GEO
Which questions can't search engines or AI answers find you for?
Search engines and AI answer engines match questions to content by meaning, not just keywords. Meaning-Flow maps the questions your audience asks against the pages you have, weights them by demand, and ranks the gaps. Paste your own lists or use the sample below; everything runs in your browser, and nothing you paste leaves it.
Coverage of demand
Loading the embedding model (about 25 MB on first visit)…
Biggest gaps
Which pages carry your demand
Demand from each covered question is credited to its closest page. Pages with little or no demand may be fine as supporting content, or may be candidates to merge.
Why this matters for AI answers
AI answer engines generally retrieve passages that are semantically close to a question before they write and cite an answer. If none of your pages is close to a question, there is little for them to retrieve, however good your site is overall. A coverage map is a first-pass check, not a guarantee: retrieval systems also weigh authority, freshness, structure and the full page text, not just titles.
How it works: questions and page titles are embedded with all-MiniLM-L6-v2, the default Sentence-BERT model in Meaning-Flow, running in your browser through Transformers.js. Coverage and gap scores follow Meaning-Flow's CoverageAnalyzer.analyze_with_embeddings: a question is covered when enough pages pass the similarity threshold, and its gap score is its volume times the share left uncovered. Source: github.com/Bodhi8/Meaning-Flow (MIT). Titles are a stand-in for full page text; the full pipeline also clusters topics and builds a topic graph.