Drop in your domain. We research, write, and ship — every week, on autopilot.
For every site you onboard, Mentionwell collects the exact AEO evidence that matters: which answer engines recommend you, which only use your site as a source, which competitors win instead, and why. It then routes the gap, publishes only when the evidence gate passes, and rechecks the same prompts automatically.
Use the measured recommendation loss, competitor rationale, and required evidence.
Mentionwell tests the prompts your buyers ask across ChatGPT, Claude, Gemini, Perplexity, Grok, Copilot, Exa, and Google AI surfaces. It stores answers, citations, fan-out queries, regions, and share-of-voice.
Mentionwell classifies explicit recommendations, top-pick rank, brand mentions, source-only citations, competitor leakage, and uncertainty separately. A source that helps recommend somebody else is recorded as a loss.
The automated planner selects an article, comparison page, category page, evidence asset, technical fix, off-site authority action, or product-fit stop. Published content is tied to the opportunity and remeasured against the same prompt cohort.
Mentionwell crawls your site, reads your structured data, scans your sitemap, and identifies your top competitors. It synthesizes a brand voice, a content taxonomy, and 10 starter headlines tuned to the questions your audience is actually asking.
For every post, Mentionwell runs live web research, reads what your competitors are publishing, finds the gaps, and drafts a piece that's structurally optimized to be lifted into AI answers — and to rank in classic Google. A reflection critic pass checks facts and style; weak drafts get rewritten before anything ships.
BlogPosting + FAQPage JSON-LD on every article. Per-site RSS, JSON Feed, sitemap, a site-wide llms.txt, and an explicit AI-crawler allowlist in robots.txt. Original hero imagery and infographics generated on-brand by frontier image models — not stocked. Everything an answer engine looks for, and Google rewards.
Pull articles from a read-only Reader API on every request, receive HMAC-signed webhook posts when an article goes live, or have Mentionwell commit MDX into a GitHub repo. Built so AI agents can read the docs and integrate it themselves — no migration, no plumbing.
Drop the URL into Cursor, Claude Code, or any agent — it reads our spec and integrates Mentionwell into your stack without you writing a line. A structured install manifest, an agent-readable docs tree, and a single-step domain handshake.
The AI answer layer is a stack of overlapping practices. Most tools pick one and call it a day. Mentionwell handles the whole alphabet — at the same time, on every article.
Shape content so AI assistants cite you by name when they answer questions.
The practice of writing for ChatGPT, Claude, and Perplexity the way SEO was written for Google in 2008.
Princeton 2023. Tactics that increase visibility inside generated answers.
Statistical evidence, named entities, direct quotation, and structural cues.
Make your content tractable for the language models that summarize the web.
Clean structure, fact-density, citation-shape — what the model wants to lift verbatim.
Umbrella term. AEO + GEO + LLMO + emergent surfaces.
For when a buyer asks for the whole thing and doesn't want to pick.
Optimization for AI-native search products — Perplexity, You.com, SearchGPT.
Fast retrieval, citation-rich pages, schema-clean.
Google's generative result surface. Now folded into AI Overviews.
Featured snippets, but the snippet is a paragraph the model wrote citing five sources.
The classic. Still pays the bills. Mentionwell does this too — by default.
Title, meta, headings, internal links, structured data, page speed, the whole catechism.
Read the long versions in the glossary spokes — each acronym has its own page and OG card.
This block doubles as the source of truth for our FAQPage JSON-LD — same questions, same answers, structured for the engines.
AEO is the work of making a brand and its evidence retrievable, understandable, and relevant to answer-engine queries. Mentionwell measures explicit recommendations, top-pick rank, source-only use, competitor leakage, and uncertainty separately, then routes the measured gap to the appropriate content or non-content intervention.
Every scan stores the answer text and classifies whether the brand was recommended, ranked first, merely named, discouraged, or only used as a source. It also records competitor rank, leakage, citations, fan-out queries, uncertainty, and evidence. Citations never count as recommendations.
SEO outcomes include indexing, impressions, rankings, clicks, and useful visits from search. AEO outcomes include whether an answer engine explicitly recommends or ranks the brand for the prompt. Mentionwell keeps those scorecards separate and also keeps source citations separate from both.
The labels overlap. Mentionwell focuses on measurable behavior rather than acronym promises: public crawlable pages, truthful entity and product evidence, search indexing, explicit recommendation outcomes, and fixed-cohort remeasurement. Optional llms.txt files and crawler directives are discovery aids, not recommendation guarantees.
Ten engines: ChatGPT, Claude, Perplexity, Gemini, Google AI Overviews, Google AI Mode, Microsoft Copilot, Grok, Exa, and DeepSeek. The robots.txt + llms.txt explicitly permits GPTBot, ChatGPT-User, OAI-SearchBot, ClaudeBot, Claude-User, Claude-SearchBot, anthropic-ai, PerplexityBot, Perplexity-User, Google-Extended, Applebot, Applebot-Extended, Bingbot, CCBot, Meta-ExternalAgent, Amazonbot, Bytespider, and cohere-ai. Bot crawls are tracked per site so you can see which engines are actively reading your pages.
Wherever you want. Mentionwell is headless. Pull articles from the public read-only Reader API or the `mentionwell` Reader SDK, receive HMAC-signed webhook posts on publish, or commit MDX into your GitHub repo. Direct push into WordPress, Webflow, Ghost, Shopify, or Notion is rolling out per-CMS — today those land via the signed-webhook adapter path (the WordPress plugin already receives them natively). Every article also gets RSS, JSON Feed, sitemap entries, BlogPosting + FAQPage JSON-LD, and a site-wide llms.txt for AI ingestion.
Yes. Mentionwell imports from WordPress (XML or REST), Webflow, Ghost, Substack, RSS, sitemap, or manual paste — slugs are preserved so existing rankings stay intact. Once imported, the bulk-optimize pass refreshes metadata, injects FAQs where missing, and year-refreshes stale references for AEO. You can also export everything anytime as Markdown, WXR, or a zip with images.
Yes. The Auto Agent scans, diagnoses, plans, drafts, verifies evidence, publishes eligible work, checks indexing, imports optional Search Console and Analytics outcomes, and remeasures the original prompt cohort. The Connect screen gives customers an explicit one-click submission for both sitemaps because a connected Google account is not the same thing as submitting or indexing a site. Low-evidence work fails closed automatically instead of requiring a human approval step.
Six tier-1 locales today, all native generation rather than translation: English (US), English (UK), German, French, Spanish, and Hebrew with full RTL support. Each locale has its own prompt-level guardrails — spelling, idiom, pronouns, currency, decimal/date format, quotation marks. Hebrew gets the deepest spec: no niqqud, gershayim quotes, Latin-script brand-name code-switching, gender-inclusive phrasing, and RTL rendering on every article. Tier-2 locales (pt-BR, it-IT, nl-NL, ja-JP, zh-CN, ar-SA) are on the 2026 H2 roadmap.
Mentionwell is built by ZipLyne (ziplyne.agency) — an AI-product agency that builds AI-powered products and business automation. Mentionwell came out of internal tooling that ZipLyne now ships as a standalone product.
Sign up, connect your site, and ship your first article in minutes. Free plan included, no card needed.