firsttechfed.com — AI Site Grade

Analysis

---

Akamai silently blocks every major AI crawler with a timeout — not a robots.txt denial, but a network-layer drop

Every AI crawler tested (GPTBot, ClaudeBot, Google-Extended, PerplexityBot, OAI-SearchBot, ChatGPT-User, anthropic-ai, Bytespider, Applebot-Extended) receives a read timeout (status 0) from the Akamai CDN edge, while a standard browser UA gets a full 200 with 122KB of content. The robots.txt is permissive (User-agent: * Disallow: /sitecore) and mentions no AI bots at all — the block is happening at the Akamai WAF/bot-manager layer, not via robots.txt directives. This means AI crawlers cannot read any page content regardless of what the site's llms.txt offers.

Crawler Access

The site runs on Akamai (evidenced by x-akamai-transformed header and server-timing). The robots.txt at https://www.firsttechfed.com/robots.txt is minimal: a single catch-all rule disallowing /sitecore and a sitemap reference. No AI-specific directives exist. However, compare_bot_access shows every AI bot UA gets a connection timeout — the Akamai edge is silently dropping these requests before any HTTP response. This is a far more aggressive block than a 403 or robots.txt denial: the crawlers never even learn the site exists behind a working edge. The llms.txt exists at a nonstandard path (/-/media/firsttech-web/llms.txt) and is well-constructed with product summaries and URL references, but it is inaccessible to the bots that need it since they cannot reach the domain at all.

Cold-Knowledge Gap

The LLM's prior knowledge about First Tech Federal Credit Union is surprisingly accurate and detailed: it knows the credit union serves tech employees (HP, Intel), has 700,000+ members, $10B+ in assets, and named specific products like the Odyssey credit card and Rewards Checking. This knowledge likely comes from pre-training data (news, Wikipedia, financial publications) rather than live crawling of the current site. The gap is that the cold model does not know about the DCU merger prominently featured on the homepage — and no external search results surface it either, suggesting this is a very recent or internal-facing announcement with no press coverage yet.

Schema Posture

The homepage and several key pages (home loans, become-a-member) carry zero JSON-LD schema. The become-a-member page has two VideoObject schemas with empty name and description fields and a uploadDate of 0001-01-01 — effectively broken. The first-tech-rewards and odyssey-rewards pages do have FAQPage schemas with well-structured Q&A, which is a positive signal. But the absence of Organization, WebSite, FinancialProduct, or LocalBusiness schema on the homepage means AI engines get no structured identity data about the credit union itself.

External Signals

External search results for First Tech Federal Credit Union are strikingly sparse. Queries for reviews, Reddit discussions, and the DCU merger returned zero results from DuckDuckGo. The site references Trustpilot reviews and mentions "Most Trusted Brand" and "Best Customer Service" honors from USA Today (2026) in its llms.txt, but these claims are not independently verifiable through search. The DNS records show integrations with Eloqua (marketing automation), Mimecast (email security), Brightcove (video), and Raymond James (investments) — a sophisticated martech stack that contrasts with the bare-bones AI visibility posture.

Findings

1. Akamai WAF silently drops all AI crawler requests with read timeouts High

   Every major AI crawler (GPTBot, ClaudeBot, Google-Extended, PerplexityBot, OAI-SearchBot, ChatGPT-User, anthropic-ai, Bytespider, Applebot-Extended) receives a read timeout from the Akamai CDN edge, while standard browsers get a full 200 response. The robots.txt is permissive and does not mention AI bots; the block occurs at the network layer, preventing crawlers from even reaching the origin.

   What to change: Reconfigure Akamai WAF or bot manager to allow AI crawler user agents (GPTBot, ClaudeBot, Google-Extended, PerplexityBot, OAI-SearchBot) to access the site, or add a dedicated allowlist for these bots.

2. llms.txt exists but is inaccessible to AI crawlers due to Akamai block High

   The site hosts a well-constructed llms.txt at a nonstandard path with product summaries and URL references, but AI crawlers cannot fetch it because the Akamai edge blocks all their requests. This renders the file useless for improving AI visibility.

   What to change: Ensure AI crawlers can access the llms.txt path by updating Akamai WAF rules to allow these bots.

3. Homepage lacks any JSON-LD schema markup High

   The homepage and several key pages (home loans, become-a-member) contain zero JSON-LD structured data. No Organization, WebSite, FinancialProduct, or LocalBusiness schema is present, depriving AI engines of structured identity and context about the credit union.

   What to change: Add JSON-LD schema for Organization, WebSite, and FinancialProduct on the homepage and key product pages.

4. Become-a-member page has broken VideoObject schemas with empty fields Medium

   The become-a-member page includes two VideoObject JSON-LD schemas with empty name and description fields and an invalid uploadDate of 0001-01-01. These broken schemas provide no value and may confuse AI parsers.

   What to change: Remove or fix the VideoObject schemas by providing valid name, description, and uploadDate values.

5. DCU merger announcement has zero external search presence Medium

   Multiple web searches for the DCU merger (e.g., 'First Tech Federal Credit Union DCU merger') returned no results on DuckDuckGo. The merger is prominently featured on the site but has no press coverage, reviews, or social mentions, limiting AI knowledge and trust signals.

   What to change: Promote the merger announcement through press releases, social media, and industry publications to generate external citations and backlinks.

6. No external review or social media presence found in search results Medium

   Searches for First Tech Federal Credit Union reviews on Trustpilot and Reddit returned zero results on DuckDuckGo. The site claims awards and positive reviews, but these are not independently verifiable through search, reducing external credibility signals for AI.

   What to change: Encourage members to leave reviews on third-party platforms and engage in social media to build external signals.

7. Robots.txt lacks AI-specific directives despite Akamai block Low

   The robots.txt file contains only a single catch-all rule disallowing /sitecore and a sitemap reference. No AI bot user agents are mentioned. While the Akamai block supersedes robots.txt, adding explicit directives would provide clarity and a fallback if the block is removed.

   What to change: Add explicit allow rules for AI crawlers (e.g., GPTBot, ClaudeBot) in robots.txt to complement the Akamai configuration.

What's working

• Well-constructed llms.txt with product summaries and URLs — The site hosts a comprehensive llms.txt file at a nonstandard path, containing summaries of key products (checking, savings, credit cards, loans) and relevant URLs. This is a strong foundation for AI visibility once crawler access is resolved.
• FAQPage schema present on rewards and credit card pages — The first-tech-rewards and odyssey-rewards pages include well-structured FAQPage JSON-LD schemas with questions and answers, which can enhance AI understanding and eligibility for rich results.
• LLM prior knowledge accurately describes the credit union — The LLM's pre-training knowledge includes accurate details about First Tech Federal Credit Union: serving tech employees, 700,000+ members, $10B+ assets, and specific products like Odyssey credit card and Rewards Checking. This indicates strong brand recognition in training data.
• Permissive robots.txt with no AI bot restrictions — The robots.txt file allows all crawlers except for the /sitecore path, which is a positive baseline. The lack of AI-specific disallow directives means the site is not intentionally blocking AI crawlers via robots.txt.
• Sitemap available with 80 URLs — The site has a sitemap at /sitemap.xml listing 80 URLs, which helps search engines discover content. This is a standard but important practice for crawlability.