Skip to content
Mentionwell
ProductFor buildersAEO PlatformPricingMentionDocs
Resources
Free AI visibility checkHow it worksIntegrationsUse casesServicesCompareChangelogAboutGlossaryBrand
Sign in Start free →
ProductFor buildersAEO PlatformPricingMentionDocs
Resources Free AI visibility checkHow it worksIntegrationsUse casesServicesCompareChangelogAboutGlossaryBrand
Sign in

Feature deep-dive

AEO Scanning

Last updated May 10, 2026

Every onboarded site gets a continuous AEO scan. We run your buyer-intent prompt corpus across up to 10 answer engines, capture every citation with claim-level evidence, trend share-of-voice, and serve up the top 5 actions per scan in a recommendations engine the Auto Agent consumes directly.

What gets scanned

  • 10 engines: ChatGPT, Claude, Perplexity, Gemini, Google AI Overviews, Google AI Mode, Microsoft Copilot, Grok, Exa, DeepSeek.
  • Up to 3 regions per scan on Scale (US, UK, EU; more on Enterprise).
  • Your prompt corpus — auto-generated from the site, editable in Prompt Explorer.

What gets captured

  • Engine answer — the full text the engine returned, with cited URLs marked.
  • Fan-out queries — sub-queries each engine searched internally.
  • Citations — domain, URL, position, snippet, source type (owned / competitor / third-party).
  • Page evidence — every cited page is scraped: schema, headings, byline, statistics, claim-level facts.
  • Share of voice — brand mentions trended by engine, locale, topic, prompt, and competitor.
  • Sentiment — positive / neutral / negative per citation.
  • Bot crawls — server-log AI bot visits trended over time.

The recommendations engine

Every scan produces a ranked top-5 actions list: prompts to defend, competitor pages to out-cite, articles to write next, schema gaps to fix, page issues to address. The Auto Agent picks from this list when it’s drafting. Operators can override via the kanban.

Questions

How are prompts generated?

From your site profile, ICP, value props, competitor list, and product surface. We auto-generate a buyer-intent corpus on day one. You can then add, edit, prioritise, mute, or import prompts in Prompt Explorer.

How often does the scan run?

Free: one-shot diagnostic. Builder: weekly on up to 15 prompts (ChatGPT + Google AI Overviews). Starter: weekly on all active prompts. Growth: daily on top-30 priority prompts, weekly on the rest. Scale: daily on every active prompt across up to 3 regions.

What's a fan-out query?

When you ask ChatGPT 'best CRM for solo founders', it internally searches several sub-queries before composing the answer (e.g. 'crm pricing', 'crm without sales team', 'pipeline tracker for one person'). Those are fan-out queries. Capturing them is half the AEO playbook — they become your H2s and FAQ entries.

Are bot crawls really useful?

Yes. Bot crawls show you which AI engines are actively reading your pages — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, etc. Rising crawl frequency is an early signal that your content is being indexed for retrieval.

See also

AEO platform overview All features Auto Agent Pricing
Live now. 14-day free trial, no card needed.Start free →Free plan: 1 article / month
Mentionwell

A blog that writes itself. Researched articles on your own domain, built to get cited by ChatGPT, Claude, Perplexity & Google AI Overviews.

Designed and Developed by ZipLyne
Product
FeaturesHow it worksAEO PlatformFree AI visibility checkAuto AgentAEO ScanningWorkspacesHeadless APIImage PipelineIndexingLanguagesImport / ExportPricing
Solutions
For buildersLovable blogBolt blogv0 blogCursor blogClaude Code blogNext.js blogReact blogIntegrationsWordPressWebflowGhostShopifyNotionUse casesFor agenciesFor SaaSCompare
Developers
Developer DocsAPI ReferenceSDKCLIMCP serversWebhooksAuthenticationQuickstart
Company
AboutThe MentionChangelogStatusContactBrandGlossary
Trust
SecurityDPASubprocessorsSLAAI DisclosurePrivacyTermsCookiesRefund PolicyAcceptable UseDMCA
© 2026 MENTIONWELL All rights reserved.