lattice.com — AI Site Grade

Analysis

Lattice's llms.txt is a rare, well-executed AI-surface — but the rest of the site's schema posture and cold-knowledge gap undermine it

Crawler Access

All major AI crawlers — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User, anthropic-ai — receive a full 200 response with identical 319KB HTML payload from Cloudflare. Bytespider is the sole bot blocked (403). The robots.txt is minimal: a catch-all User-agent: * with only four disallow rules (/wip/, /system/, /explore, /*page*page) and no AI-bot-specific directives. The site runs on Cloudflare with HSTS preload and a frame-ancestors 'self' CSP, but no X-Content-Type-Options or Referrer-Policy headers.

llms.txt and Content Surface

Lattice ships a comprehensive llms.txt (9,315 bytes) — a rare maturity signal. It lists 20+ product pages with human-written summaries covering the platform, AI Agent, performance, engagement, compensation, and goals modules. The homepage itself is text-rich (785 words) with an H1 of "People + AI: Succeeding Together" and clear product subheadings. However, no sub-page beyond the homepage carries any JSON-LD schema — the /platform, /ai, /about, /pricing, /compare, /customers, and /platform/ai-agent pages all return zero schema types. The homepage has Organization and WebSite schema with founding date, social profiles, and contact points, but the Organization schema omits foundingLocation (San Francisco) and the sameAs list excludes the community Facebook group that appears in the footer.

Cold-Knowledge Gap

The LLM's prior knowledge about Lattice contains three significant inaccuracies. It states the Thoma Bravo acquisition occurred in 2022 (the site's blog and press releases confirm it was 2024). It claims pricing is not publicly listed and starts at $11–15/seat — the site's /pricing page publicly lists $11/seat for Foundations and unbundled modules at $8/seat for Performance or Goals & OKRs. It describes Lattice as serving "mid-market to enterprise" with a "strong base in tech" — the site prominently features customers from manufacturing (Loop Earplugs), hospitality (USAirtours), food & beverage (UNREAL Snacks, Coca-Cola), and credit unions (Vantage West). The cold knowledge also mentions "employee records" as a product feature, which does not appear anywhere on the live site.

Schema and Answer-Signal Posture

The homepage has Organization + WebSite schema, but the Organization block uses a generic description field rather than the site's actual meta description. The /pricing page has an FAQ section with 5 questions but no FAQ schema — a missed opportunity for AI answer extraction. The /ai page has a 6-question FAQ section with substantive answers (including a notable disclaimer that "Lattice does not use AI to write performance reviews for employees") but again no FAQ schema. The /compare page lists 9 competitor comparisons (Culture Amp, Leapsome, 15Five, PerformYard, ADP, Paylocity, BambooHR, HiBob, Rippling) but each comparison page is a thin shell with no structured comparison data or table schema.

External Signals and Surprising Findings

The DNS TXT records reveal an anthropic-domain-verification token and an openai-domain-verification token — confirming Lattice has proactively verified its domain with both Anthropic and OpenAI for AI crawler trust. The blog shows product updates through April 2026, indicating active development and fresh content. The /customers page lists 20+ case studies across diverse industries, but the /compare pages for individual competitors (e.g., /compare/culture-amp) were not found in the sitemap — the /compare page is a hub with no deep-linked competitor pages discoverable to crawlers. The site uses a Webflow-like stack (headers reveal x-wf-page-id, x-wf-region, x-lambda-id), which is a server-rendered setup — no JS-rendering risk for AI crawlers.

Findings

1. Pricing page FAQ section lacks FAQ schema Medium

   The /pricing page contains 5 FAQ questions but no FAQ schema markup, missing an opportunity for AI answer extraction.

   What to change: Add FAQ schema to the FAQ section on the /pricing page.

2. AI page FAQ section lacks FAQ schema Medium

   The /ai page has 6 FAQ questions with substantive answers but no FAQ schema markup.

   What to change: Add FAQ schema to the FAQ section on the /ai page.

3. No JSON-LD schema on any subpage beyond homepage High

   Pages like /platform, /ai, /about, /pricing, /compare, /customers, and /platform/ai-agent return zero schema types, reducing AI visibility.

   What to change: Add appropriate JSON-LD schema (e.g., Product, FAQPage, SoftwareApplication) to all key subpages.

4. LLM cold knowledge contains three significant inaccuracies High

   LLM prior knowledge incorrectly states the Thoma Bravo acquisition occurred in 2022 (actual: 2024), pricing is not public (actual: $11/seat on /pricing), and target industries are mid-market/tech (actual: includes manufacturing, hospitality, food & beverage, credit unions).

   What to change: Publish a press release or blog post about the 2024 acquisition, ensure pricing page is clearly visible, and highlight diverse industry case studies on the homepage.

5. Homepage Organization schema omits founding location and community Facebook Low

   The Organization schema on the homepage lacks foundingLocation (San Francisco) and the sameAs list excludes the community Facebook group linked in the footer.

   What to change: Add foundingLocation and include the community Facebook URL in sameAs.

6. Individual competitor comparison pages not in sitemap Medium

   The /compare page is a hub listing 9 competitors, but individual comparison pages (e.g., /compare/culture-amp) are not found in the sitemap, limiting crawler discovery.

   What to change: Add individual competitor comparison pages to the sitemap and ensure they are linked from the hub page.

7. Bytespider bot blocked with 403 Low

   Bytespider receives a 403 response, preventing Baidu's AI crawler from accessing the site.

   What to change: Consider allowing Bytespider if Baidu traffic is relevant.

8. Missing X-Content-Type-Options and Referrer-Policy headers Low

   The site lacks X-Content-Type-Options and Referrer-Policy headers, which are minor security and SEO signals.

   What to change: Add X-Content-Type-Options: nosniff and Referrer-Policy: strict-origin-when-cross-origin headers.

What's working

• Comprehensive llms.txt with 20+ product pages — Lattice ships a 9,315-byte llms.txt listing 20+ product pages with human-written summaries, a rare maturity signal for AI visibility.
• Domain verified with Anthropic and OpenAI — DNS TXT records show anthropic-domain-verification and openai-domain-verification tokens, indicating proactive AI crawler trust setup.
• All major AI crawlers allowed with full HTML access — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User, and anthropic-ai all receive 200 responses with identical HTML, ensuring AI crawlers can index the site.
• Server-rendered Webflow-like stack avoids JS rendering issues — The site uses a server-rendered stack (Webflow-like), so AI crawlers receive full HTML content without JavaScript rendering risks.
• Homepage has Organization and WebSite schema — The homepage includes Organization and WebSite JSON-LD schema with founding date, social profiles, and contact points.
• Blog shows product updates through April 2026 — The blog features recent product updates, indicating active development and fresh content for AI crawlers.
• 20+ customer case studies across diverse industries — The /customers page lists case studies from manufacturing, hospitality, food & beverage, and credit unions, demonstrating broad market appeal.