# How ChatGPT, Claude, and Bing actually find Retain Inc

This note is for operators after launch. It is **not** marketing copy for the public homepage.

There is no meta tag, `llms.txt` file, or MCP server that guarantees Retain Inc will appear when someone types a question into consumer ChatGPT or Claude. Those products search and browse the public web (ChatGPT Search leans on Bing plus OpenAI crawlers). Files in this repo make the site **fetchable and citable**. Ranking still depends on being live on HTTPS, indexed by search engines, and mentioned as a real brand.

MCP in `/mcp/retaininc-mcp/` helps Cursor or Claude Desktop read this **repository**. It does not submit the site to ChatGPT Search.

## What this repo already does

- `robots.txt` explicitly **Allow: /** for GPTBot, ChatGPT-User, OAI-SearchBot, ClaudeBot, anthropic-ai, Claude-User, Applebot-Extended, Googlebot, Google-Extended, Bingbot, PerplexityBot, Bytespider, and `User-agent: *`.
- No `noindex` on home, services, email, CRM, or the resources hub.
- Sitemap index at https://retaininc.com/sitemap.xml (also listed in robots.txt).
- `llms.txt` / `llms-full.txt` and curated `/ai/*.md` so crawlers that prefer plain text get accurate facts.
- `feed.xml` for feed-oriented crawlers.
- Organization + WebSite JSON-LD on the homepage **without** a fake `SearchAction` (this site has no on-site search box).

## What you must do after the site is live on HTTPS

1. Confirm https://retaininc.com/robots.txt , https://retaininc.com/llms.txt , and https://retaininc.com/sitemap.xml load without login or blocking headers.
2. Add the property in [Google Search Console](https://search.google.com/search-console) and submit `https://retaininc.com/sitemap.xml`.
3. Add the site in [Bing Webmaster Tools](https://www.bing.com/webmasters) and submit the same sitemap. ChatGPT Search relies heavily on Bing’s index; skipping Bing is the most common miss.
4. Use brand mentions that are true: directory listings, partner pages, LinkedIn, and earned coverage that name **Retain Inc** and **retaininc.com**. Models and search stacks use that corroboration.
5. Wait. Crawler and index updates are not instant (OpenAI has documented on the order of ~24 hours for robots.txt changes; Bing/Google indexing of a new site is often much longer).

## IndexNow

Do not invent an IndexNow key file. After you create a real key in Bing Webmaster Tools, place `{key}.txt` at the site root with that key as the only contents, then ping IndexNow when important URLs change. Until then, sitemap submission is the discovery path.

## Optional checks

- `curl -I https://retaininc.com/` should be 200, `index,follow` (or no robots meta), HTTPS.
- Fetch as Google / URL inspection in GSC and Bing for `/`, `/services/email-marketing.html`, `/services/crm-marketing.html`, `/resources/`.
- Confirm CDN/WAF is not blocking OpenAI, Anthropic, or Bing user-agents.