AI Site Grade
amplitude.com — AI Site Grade
Amplitude's homepage JSON-LD points to a rebrand page as canonical, contradicting the AI Visibility product it sells.
Amplitude has best-in-class AI crawler access and a sophisticated llms.txt ecosystem, but its homepage lacks core schema and its LLM cold knowledge is 18 months stale, missing the AI Visibility product entirely.
- Findings
- 8
- Evidence checks
- 20
- Completed
- 30 May 2026
Analysis
Amplitude.com — AI-Visibility Audit
The site sells an "AI Visibility" product to help brands track how LLMs describe them, yet the homepage's own JSON-LD schema references a rebrand page (/a-rebrand) as its canonical URL — a structural contradiction that undermines the very credibility the product promises.
Crawler Access
All major AI crawlers — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, Bytespider, Applebot-Extended, anthropic-ai, ChatGPT-User, Perplexity-User — receive a 200 with full HTML content (522–542 KB, matching the browser baseline). No UA-based blocking, no JS shell, no Cloudflare challenge. The robots.txt explicitly allows every listed AI bot to crawl / while only disallowing /*sanity_test* and /*.md$ paths. The site runs on Vercel (Next.js) behind nginx with HSTS and a moderate CSP. This is a best-in-class technical AI-crawl posture.
LLM-First Infrastructure
Amplitude operates a three-tier llms.txt ecosystem — a rarity. /llms.txt (2764 bytes) serves as an AI coding-agent primer for the @amplitude/wizard CLI. /llms-full.txt (18 KB) is a full agent reference with NDJSON schemas, exit codes, and MCP server config. /sitemap-llms.txt (810 KB) is a curated, categorized index of every public page on the domain with human-readable descriptions. This is the most sophisticated llms.txt deployment observed on any commercial SaaS domain.
Cold-Knowledge Gap
The LLM prior knows Amplitude as a product analytics platform that pioneered behavioral analytics, went public via direct listing in 2021 (AMPL), and faced post-IPO headwinds including 2023 layoffs. The prior mentions Amplitude Experiment, Recommend, and CDP — but does not mention AI Visibility, AI Agents, AI Feedback, or the MCP server at all. The site now leads with "The AI analytics platform" and positions AI Visibility as a flagship product. The model's cold knowledge is ~18 months stale — it describes a pre-AI-rebrand company that no longer matches the homepage messaging.
Schema Posture
The homepage carries only a WebPage schema — no Organization, SoftwareApplication, or Product schema on the primary marketing page. The WebPage schema's url field points to https://amplitude.com/a-rebrand rather than /. The /amplitude-analytics product page does include a FAQPage schema with 5 well-formed Q&A entries. The /ai-visibility page also carries a FAQPage schema with 9 questions. The /docs page has a SoftwareApplication schema. But the homepage — the page most LLMs retrieve first — lacks structured data describing what Amplitude actually is (a software platform, an organization, a product).
External Signals
The /amplitude-vs-mixpanel comparison page redirects to info.amplitude.com (a Marketo landing page) with only 102 words of visible text and no schema. The /amplitude-vs-pendo page redirects to /sales-contact. Both comparison pages are effectively dead ends for AI crawlers — thin content with no structured data, undermining the competitive-intelligence content an LLM would cite when comparing tools. The blog is active (multiple posts in May 2026) and heavily focused on AI agents, CLI tools, and Global Agent — content that aligns with the new AI positioning.
Findings
Homepage JSON-LD canonical URL points to /a-rebrand High
The homepage's WebPage schema sets its canonical URL to https://amplitude.com/a-rebrand instead of the actual homepage. This misdirects LLMs and undermines the credibility of the AI Visibility product.
What to change: Update the homepage JSON-LD to set the canonical URL to https://amplitude.com and add Organization and SoftwareApplication schema.
Homepage lacks Organization and SoftwareApplication schema High
The homepage only has a WebPage schema. It does not include Organization, SoftwareApplication, or Product schema, which are critical for LLMs to understand the brand and product.
What to change: Add Organization and SoftwareApplication schema to the homepage with accurate name, description, and URL.
LLM cold knowledge omits AI Visibility product High
The LLM prior describes Amplitude as a product analytics platform and does not mention AI Visibility, AI Agents, or the MCP server. The model's knowledge is approximately 18 months stale, misrepresenting the current product positioning.
What to change: Publish structured data and content that explicitly describes the AI Visibility product, and consider submitting updated information to LLM knowledge bases.
Comparison pages have thin content and no schema Medium
The /amplitude-vs-mixpanel page redirects to a Marketo landing page with only 102 words and no schema. The /amplitude-vs-pendo page redirects to /sales-contact. These pages provide little value for LLMs seeking competitive comparisons.
What to change: Create substantive comparison pages with detailed feature tables, customer quotes, and FAQPage schema.
/llms-full.txt returns empty content Medium
The /llms-full.txt URL returns a 200 status but with 0 bytes of content. This breaks the expected fallback for LLMs that request the full file.
What to change: Populate /llms-full.txt with the full agent reference content as intended.
/sitemap-llms.txt returns empty content Medium
The /sitemap-llms.txt URL returns a 200 status but with 0 bytes. This curated index is missing, reducing discoverability.
What to change: Populate /sitemap-llms.txt with the categorized page index.
Docs sitemap XML returns empty content Medium
The /docs/sitemap-docs.xml URL returns a 200 status but with 0 bytes, preventing crawlers from discovering documentation pages.
What to change: Ensure the docs sitemap is populated with all documentation URLs.
Root sitemap XML returns empty content Medium
The /sitemap-root.xml URL returns a 200 status but with 0 bytes, preventing crawlers from discovering the site structure.
What to change: Populate the root sitemap with all public URLs.
What's working
- All major AI crawlers receive full HTML content — Every major AI crawler (GPTBot, ClaudeBot, PerplexityBot, etc.) gets a 200 response with full HTML, no blocking, no JS shell. This is a best-in-class technical AI-crawl posture.
- Three-tier llms.txt ecosystem with /llms.txt, /llms-full.txt, /sitemap-llms.txt — Amplitude operates a rare three-tier llms.txt deployment: /llms.txt serves as an AI coding-agent primer, /llms-full.txt is a full agent reference, and /sitemap-llms.txt is a curated page index. This is the most sophisticated deployment observed on any commercial SaaS domain.
- FAQPage schema on /amplitude-analytics and /ai-visibility pages — The /amplitude-analytics page has a FAQPage schema with 5 Q&A entries, and the /ai-visibility page has 9 questions. These help LLMs extract structured product information.
- SoftwareApplication schema on /docs page — The /docs page includes a SoftwareApplication schema, which helps LLMs identify the documentation as belonging to a software product.
- Blog heavily focused on AI agents and CLI tools — The blog contains multiple posts in May 2026 about AI agents, CLI tools, and Global Agent, aligning with the new AI positioning and providing fresh content for LLMs.
Track amplitude.com across AI search
This is one snapshot. Open the interactive report to inspect evidence, or grade another site free.