runa.io — AI Site Grade

Analysis

Runa's AI Crawler Blockade Contradicts Its OpenAI Domain Verification

Runa's DNS includes an openai-domain-verification TXT record, signaling intent to be crawled by OpenAI — yet every OpenAI crawler (GPTBot, OAI-SearchBot, ChatGPT-User) receives a 403 block from Cloudflare, identical to ClaudeBot, PerplexityBot, Bytespider, and anthropic-ai. Only Google-Extended and Applebot-Extended pass through to the full 374KB page alongside a standard browser.

Crawler Access

The robots.txt at runa.io/robots.txt contains a single catch-all User-agent: * rule with no AI-bot-specific directives whatsoever. No GPTBot, no ClaudeBot, no PerplexityBot — nothing. The Allow: / line is present, but Cloudflare's WAF overrides it at the network layer. compare_bot_access confirmed that 8 of 11 tested AI crawlers are blocked with 403 and a 25-byte response. The site runs on Cloudflare (A records point to 104.18.12.185/13.185, nginx server header) with WP Engine hosting (X-Cache-Group headers, wpenginepowered.com canonical leakage). The llms.txt returns a 404. The research page at /research/ is a JS-dependent shell — it loads 20 words of visible text and a "Loading resources..." spinner that never resolves on a plain GET.

Schema Posture

Every page carries the same minimal WebPage + BreadcrumbList + WebSite schema triad. No Organization schema with logo, social profiles, or founding date. No Product or SoftwareApplication schema for the payout API. No FAQPage schema despite the /products/network/ page containing a Q&A block ("How does the Runa Network streamline global payouts...") rendered as plain text. The WebSite schema's description field is empty string on every page. The /developers/ page's canonical URL points to runaio.wpenginepowered.com/developers/ — a staging/hosting domain leak.

Cold-Knowledge Gap

The LLM knows Runa as a London-founded 2016 payout platform competing with Tremendous and Trolley, serving gig economy and loyalty clients. The site itself positions far more aggressively: "The World Gets Paid Through Runa," "5,000+ merchant gift card endpoints," "190+ countries," "99.999% uptime," "sub-second order & delivery." The cold model knows nothing about the Runa Prepaid product, the Runa Shop merchant network, the Runa Assure compliance layer, or the Decoding Payouts research study — all prominently featured on the live site. The model also cites "over 200 countries" while the site consistently says 190+.

External Signals

The press release at /press-release/slow-payouts-drive-worker-attrition/ was distributed via BusinessWire and cites named investors (Element Ventures, CommerzVentures, Clocktower Ventures, Volution Capital, 13Books Capital, SAP). The site references case studies with Vodafone, Pluxee, YuLife, VoucherCodes, and Azteco — but web searches for "runa.io" and "Runa payouts reviews" returned zero organic results, suggesting minimal off-domain citation footprint. The DNS TXT records reveal integrations with HubSpot, Zendesk, Salesforce, Mandrill, Mixpanel, Notion, Canva, Postman, and Atlassian — a rich tech stack that the site's schema never surfaces.

Canonical and Redirect Issues

The /customers/case-study-vodafone URL redirects to /case-studies/vodafone/ (301), but the sitemap still lists the old path. The /developers/ page's canonical is set to runaio.wpenginepowered.com/developers/ — a WP Engine staging domain that should never appear in production canonicals. The dateModified field on the homepage reads 2026-04-17, which is a future date at time of writing, indicating a WordPress plugin or theme is injecting incorrect timestamps.

Findings

1. OpenAI crawlers blocked by Cloudflare despite domain verification High

   Runa's DNS includes an openai-domain-verification TXT record, signaling intent to be crawled by OpenAI, but GPTBot, OAI-SearchBot, and ChatGPT-User all receive 403 blocks from Cloudflare. Only Google-Extended and Applebot-Extended pass through.

   What to change: Update Cloudflare WAF rules to allow GPTBot, OAI-SearchBot, and ChatGPT-User based on their user-agent strings.

2. Robots.txt lacks AI-bot directives Medium

   The robots.txt file contains only a catch-all User-agent: * rule with no specific directives for GPTBot, ClaudeBot, PerplexityBot, or other AI crawlers. Cloudflare's WAF overrides the Allow: / line.

   What to change: Add explicit Allow or Disallow rules for each major AI crawler in robots.txt to align with the intended access policy.

3. llms.txt returns 404 Medium

   The standard llms.txt file, which helps AI crawlers discover key content, is not present and returns a 404 error.

   What to change: Create an llms.txt file listing important pages such as /products/network, /developers, and /why-runa.

4. Research page renders as empty JS shell High

   The /research/ page loads only 20 words of visible text and a 'Loading resources...' spinner that never resolves on a plain GET, making its content invisible to crawlers that do not execute JavaScript.

   What to change: Ensure the research page content is server-side rendered or statically generated so it is accessible without JavaScript.

5. WebSite schema description is empty on every page Medium

   The WebSite schema's description field is an empty string on all pages, missing a critical opportunity to communicate the site's purpose to search engines and AI crawlers.

   What to change: Populate the WebSite schema description with a concise summary of Runa's platform, e.g., 'Runa is a digital payouts platform enabling businesses to send bulk payments globally.'

6. No Organization schema on any page Medium

   The site lacks Organization schema with logo, social profiles, founding date, or other brand signals, which limits entity recognition by AI models.

   What to change: Add Organization schema to the homepage and key pages, including logo URL, sameAs links to social profiles, and founding date.

7. No Product or SoftwareApplication schema for payout API Medium

   The site prominently features a payout API and network, but no Product or SoftwareApplication schema is present to describe these offerings to search engines and AI crawlers.

   What to change: Add SoftwareApplication schema to the /developers page and Product schema to /products/network with appropriate properties.

8. FAQPage schema missing from Q&A content Low

   The /products/network page contains a Q&A block rendered as plain text, but no FAQPage schema is used, missing an opportunity for rich results and AI extraction.

   What to change: Wrap the Q&A content in FAQPage schema markup.

9. Developers page canonical points to WP Engine staging domain High

   The /developers/ page's canonical URL is set to runaio.wpenginepowered.com/developers/, leaking a staging/hosting domain into production and confusing crawlers.

   What to change: Update the canonical URL on the /developers page to https://runa.io/developers.

10. Sitemap lists old URL that redirects Low

    The sitemap includes /customers/case-study-vodafone, which 301 redirects to /case-studies/vodafone/, wasting crawl budget and potentially confusing crawlers.

    What to change: Update the sitemap to use the final redirect target URL /case-studies/vodafone/.

11. Homepage dateModified set to future date Medium

    The homepage's dateModified field reads 2026-04-17, a future date, likely due to a WordPress plugin or theme injecting incorrect timestamps, which can confuse crawlers and reduce trust.

    What to change: Correct the dateModified field to the actual last-modified date or remove it if dynamically generated incorrectly.

12. No organic search results for Runa brand terms High

    Web searches for 'runa.io', 'Runa payouts platform reviews', and 'runa.io payout API gift cards' returned zero results, indicating minimal off-domain citation footprint and low external signals.

    What to change: Invest in content marketing, backlink building, and listing on review platforms like G2 and Capterra to increase organic visibility.

What's working

• OpenAI domain verification TXT record is present — The DNS includes an openai-domain-verification TXT record, indicating intent to be crawled by OpenAI and enabling potential future access.
• Google-Extended and Applebot-Extended are allowed — Google-Extended and Applebot-Extended pass through Cloudflare and can access the full page content, ensuring visibility to Google's AI and Apple's services.
• Basic WebPage, BreadcrumbList, and WebSite schema present — Every page includes a minimal schema triad (WebPage, BreadcrumbList, WebSite), providing foundational structured data for search engines.
• Press release with named investors distributed via BusinessWire — The press release at /press-release/slow-payouts-drive-worker-attrition/ cites named investors and was distributed via BusinessWire, providing some external credibility signals.
• Case studies with Vodafone and other brands — The site features case studies with Vodafone, Pluxee, YuLife, VoucherCodes, and Azteco, providing social proof and potential backlink opportunities.
• DNS TXT records reveal integrations with major platforms — DNS TXT records show integrations with HubSpot, Zendesk, Salesforce, Mandrill, Mixpanel, Notion, Canva, Postman, and Atlassian, indicating a sophisticated tech stack that could be surfaced via schema.
• Blog and developer pages are accessible and content-rich — The /blog and /developers pages return 200 with substantial content (984 and 168 words respectively), providing valuable information for crawlers.
• Sitemap is present with 80 URLs — The sitemap at /sitemap.xml contains 80 URLs, providing a comprehensive list of pages for crawlers to discover.