asuresoftware.com — AI Site Grade

Analysis

The Entire Site Is Invisible to Browsers but Fully Open to AI Crawlers

Every page on asuresoftware.com returns HTTP 403 to browser user-agents while serving full 200-byte responses to GPTBot, Google-Extended, PerplexityBot, OAI-SearchBot, and anthropic-ai. The site is a JS-rendered WordPress deployment on WP Engine behind Cloudflare, and the Cloudflare WAF is configured to block all non-bot traffic — including the analyst's browser — while letting AI crawlers pass freely. This creates a bizarre inversion: human visitors see only "403 Forbidden nginx," while AI models ingest the full marketing content.

Crawler Access

The robots.txt at https://www.asuresoftware.com/robots.txt contains only a single User-agent: * rule disallowing /wp-admin/ and allowing /wp-admin/admin-ajax.php, with a crawl-delay of 10 seconds. No AI-specific bot rules exist — no mention of GPTBot, ClaudeBot, Google-Extended, or any other AI crawler. The llms.txt at https://asuresoftware.com/llms.txt returns HTTP 403 to all requesters, effectively absent. The sitemap at https://www.asuresoftware.com/wp-sitemap.xml is accessible to AI bots and reveals approximately 1,500+ blog posts and 80+ pages, but sub-sitemaps (posts, pages, categories) are blocked for browser UAs. ClaudeBot received a 429 rate-limit during testing, suggesting inconsistent bot treatment.

Content & Schema

The homepage (captured via Wayback Machine, since live browser requests are blocked) positions Asure as a payroll, tax, and HR outsourcing provider serving 100,000+ businesses with two service models: self-service via AsureCentral and managed via AsureWorks. The site prominently features Luna AI as an embedded assistant, SHRM-certified professionals, SOC 1 & SOC 2 Type II certification, and SEC-regulated public company status (NASDAQ: ASUR). No JSON-LD schema is present on any page — no Organization, Product, FAQPage, or BreadcrumbList markup was detected. The heading structure is well-organized with clear H1-H3 hierarchy, and the content includes comparison language ("Run payroll yourself or have Asure's experts manage it"), tables, and list-based feature grids that answer-format signals would reward.

Cold-Knowledge Gap

The LLM's prior knowledge of Asure Software is moderately accurate but outdated and incomplete. It correctly identifies the company as a cloud-based HCM provider founded in 1985, publicly traded under ASUR, and serving SMBs. However, it makes no mention of the AsureWorks vs. AsureCentral dual-model structure, the Luna AI assistant, the Employer Growth Standard framework, or the AsurePay earned-wage-access product — all of which are central to the current site messaging. The prior also references "occasional issues with customer support responsiveness" from review platforms, which the site itself does not address.

External Signals

The DNS TXT records reveal domain verifications for Anthropic, Apple, Atlassian, BrowserStack, Google, Slack, and Smartsheet — indicating active integrations or partnerships. The anthropic-domain-verification token confirms Asure has explicitly verified ownership for Claude's training pipeline. No external reviews, press mentions, or Reddit threads were surfaced during investigation, suggesting limited off-domain footprint that AI engines would draw upon for brand descriptions.

Findings

1. All pages return HTTP 403 to browser user-agents High

   Every page on asuresoftware.com returns a 403 Forbidden error when accessed by a standard browser, blocking human visitors entirely while AI crawlers receive full 200 responses.

   What to change: Remove the Cloudflare WAF rule that blocks non-bot user-agents, or configure it to allow human browsers while maintaining security.

2. No AI-specific rules in robots.txt Medium

   The robots.txt file contains only a generic User-agent: * rule disallowing /wp-admin/ and does not mention GPTBot, ClaudeBot, Google-Extended, or any other AI crawler.

   What to change: Add explicit directives for AI crawlers (e.g., GPTBot, ClaudeBot, Google-Extended) to control access to specific content.

3. llms.txt returns HTTP 403 Medium

   The llms.txt file at https://asuresoftware.com/llms.txt is inaccessible, returning a 403 Forbidden response, which prevents AI crawlers from discovering a curated content index.

   What to change: Create and serve a valid llms.txt file listing key pages for AI crawlers.

4. GPTBot receives HTTP 429 rate limit Medium

   GPTBot is rate-limited (429) when accessing the homepage and sitemap, while other AI bots like OAI-SearchBot and PerplexityBot receive 200 responses.

   What to change: Remove or adjust rate-limiting rules for GPTBot to ensure consistent access.

5. No JSON-LD structured data on any page High

   No JSON-LD schema markup (Organization, Product, FAQPage, BreadcrumbList) was detected on the homepage or any other page, reducing the site's ability to provide rich snippets to search engines and AI crawlers.

   What to change: Add JSON-LD structured data for Organization, Product, FAQPage, and BreadcrumbList across all relevant pages.

6. LLM prior knowledge missing key products and features Medium

   The LLM's prior knowledge of Asure Software lacks mention of the AsureWorks vs. AsureCentral dual-model, Luna AI assistant, Employer Growth Standard framework, and AsurePay earned-wage-access product, all central to current site messaging.

   What to change: Ensure key products and features are prominently mentioned in crawlable content and structured data to improve AI knowledge.

7. Sitemap sub-files blocked for browser user-agents Medium

   Sub-sitemaps (posts, pages, categories) return 403 for browser user-agents, though they are accessible to AI bots like OAI-SearchBot and PerplexityBot.

   What to change: Allow browser user-agents to access sitemap sub-files to improve search engine indexing.

8. Limited external mentions and reviews Medium

   Web searches for Asure Software returned zero results for reviews, press mentions, or social media discussions, indicating a weak off-domain footprint that AI engines rely on for brand descriptions.

   What to change: Encourage customer reviews, press coverage, and social media activity to build external signals.

9. ClaudeBot receives HTTP 429 rate limit Low

   ClaudeBot (anthropic-ai) received a 429 rate-limit response during testing, suggesting inconsistent bot treatment despite domain verification for Anthropic.

   What to change: Ensure ClaudeBot is not rate-limited, given the domain verification token for Anthropic.

What's working

• AI crawlers receive full 200 responses with rich content — OAI-SearchBot, PerplexityBot, Google-Extended, and anthropic-ai all receive full HTML content (200-250KB) from the homepage and key pages, enabling AI models to ingest marketing material.
• Domain verified for Anthropic via DNS TXT record — The DNS TXT records include an anthropic-domain-verification token, confirming Asure has explicitly verified ownership for Claude's training pipeline.
• Well-organized heading hierarchy with clear H1-H3 structure — The homepage and key pages use a clear heading hierarchy (H1-H3) with descriptive titles, which helps AI crawlers parse content structure.
• Comparison language and feature grids provide answer-format signals — Pages include comparison tables, list-based feature grids, and decision-oriented content (e.g., "Run payroll yourself or have Asure's experts manage it") that align with AI answer-format preferences.
• Sitemap and sub-sitemaps accessible to AI crawlers — The main sitemap and sub-sitemaps (posts, pages, categories) are fully accessible to OAI-SearchBot and PerplexityBot, providing a complete content inventory to AI crawlers.
• DNS TXT records show verifications for multiple platforms — Domain verification tokens for Apple, Atlassian, BrowserStack, Google, Slack, and Smartsheet indicate active integrations or partnerships that may boost credibility signals.