AI Site Grade
zilch.com — AI Site Grade
Zilch.com's AI crawler access is fully open but lacks explicit signposting, while LLM knowledge lags behind the site's evolved positioning and schema coverage is sparse.
Zilch.com permits all AI crawlers but has no robots.txt directives for them, no /llms.txt, outdated LLM knowledge, and minimal schema on key pages.
- Findings
- 9
- Evidence checks
- 23
- Completed
- 30 May 2026
Analysis
Zilch.com — AI-Visibility Audit
The site's robots.txt contains zero AI-bot directives — no GPTBot, no ClaudeBot, no Google-Extended, no PerplexityBot — yet every single AI crawler tested receives a full 200 with the same 187KB payload as a browser, making this one of the most permissively accessible fintech sites from a crawler standpoint, but also one of the least explicitly signposted.
Crawler Access
All 11 AI user-agents tested (GPTBot, ClaudeBot, Google-Extended, PerplexityBot, OAI-SearchBot, ChatGPT-User, anthropic-ai, Applebot-Extended, Perplexity-User, Bytespider, ChatGPT-User) return HTTP 200 with identical byte size (187,663) to the browser baseline. The sole exception is Bytespider (ByteDance), which gets a 502 from Cloudflare — likely a WAF block, not a deliberate choice. The site runs on Cloudflare (CDN/WAF) with WP Engine hosting, serving over AWS DNS. No JS-rendering risk: the homepage delivers ~1,200 words of visible text on a plain GET. The /llms.txt URL does not exist — it resolves to the homepage (200, same content), meaning no AI-friendly content map is published.
Cold-Knowledge Gap
The LLM's prior knowledge describes Zilch as a "UK-based fintech" using a "Mastercard virtual card" with "pay in 4" and a "$110 million Series C in 2021" valuing it at "over $2 billion." The actual site tells a materially different story: Zilch now claims "over 6 million customers" (not 5 million), offers a tiered subscription model (Standard free, Plus at £3.99/month, with Extra/Pro/Ultimate/Business "coming soon"), has partnered with Arsenal F.C., and is profitable across its product suite. The cold model still cites the 2021 valuation and mentions FCA scrutiny from 2023 — neither of which appears on the current site, which instead highlights FCA authorisation (FRN 843421) as a trust signal. The model also says "Mastercard" while the UK site now references Visa acceptance ("37 million+ merchants") and the US site references Cross River Bank partnership. This is a significant positioning drift the model has not caught.
Schema Posture
The UK homepage carries a WebSite + WebPage + Person/Organization graph but no Product schema, no FAQPage on pages that clearly contain FAQ content, and no FinancialService typing on the UK site (the US site does use FinancialService on its homepage). The corporate page (/corporate/home/) is the only page found with a proper FAQPage schema — containing three well-structured Q&A entries about how Zilch works, costs, and credit card comparison. The blog ("The Green") and product pages (/uk/zilch-plus/, /uk/zilch-rewards/) have zero schema beyond the site-wide graph. The dateModified on the UK homepage reads "2026-04-08" — a future date, likely a CMS bug that could confuse freshness signals.
External Signals
The site prominently embeds Trustpilot reviews ("4.6 / 5.0") and cites an NPS of 75 (as of April 2025). The retailers page lists major partners (Iceland, Morrisons, Awin, eBay) and claims 5,000+ retailers, 2.5 billion yearly sales, and 40%+ higher AOV. The Arsenal F.C. partnership is a recent high-signal brand move. The US site is notably thinner — ~390 words, no blog, no FAQ, and a waitlist-driven signup flow — suggesting the US expansion is still early-stage. The FAQ pages live on a separate subdomain structure (/faq/) that is not linked from the UK sitemap, creating a potential crawl gap for AI engines that rely on sitemap discovery.
Findings
No AI-bot directives in robots.txt Medium
The robots.txt file contains no directives for GPTBot, ClaudeBot, Google-Extended, or any other AI crawler, leaving access unmanaged.
What to change: Add explicit allow or disallow rules for known AI crawlers to control how content is consumed.
Missing /llms.txt file Medium
The /llms.txt URL resolves to the homepage instead of serving an AI-friendly content map, missing an opportunity to guide LLM consumption.
What to change: Publish a /llms.txt file listing key pages and summaries for AI crawlers.
LLM knowledge lags behind current site positioning High
LLM prior knowledge describes Zilch with outdated facts (2021 valuation, Mastercard, 5 million customers) while the site now features 6 million customers, Visa acceptance, tiered subscriptions, and Arsenal partnership.
What to change: Publish structured data and authoritative content to help LLMs refresh their knowledge, such as a company fact sheet with current metrics.
FAQ pages lack FAQPage schema Medium
Pages like /faq/how-does-zilch-work/ contain FAQ content but do not use FAQPage schema, reducing their chance of appearing in rich results.
What to change: Add FAQPage structured data to all FAQ pages.
Product and blog pages lack schema markup Medium
Pages like /uk/zilch-plus/, /uk/zilch-rewards/, and /uk/thegreen/ have no schema beyond the site-wide graph, missing opportunities for rich snippets.
What to change: Add Product, Article, or FinancialService schema to relevant pages.
UK homepage has future dateModified Low
The UK homepage's dateModified is set to 2026-04-08, a future date that may confuse search engines about content freshness.
What to change: Correct the dateModified to the actual last update date.
FAQ subdomain not linked from UK sitemap Medium
FAQ pages live on a separate subdomain structure (/faq/) that is not included in the UK sitemap, creating a crawl gap for AI engines.
What to change: Include FAQ URLs in the UK sitemap or ensure cross-linking from main pages.
US site is thin with limited content Medium
The US site has only ~390 words, no blog, no FAQ, and a waitlist-driven signup, offering little for AI crawlers to index.
What to change: Expand US site content with product details, FAQ, and blog to improve AI visibility.
Bytespider blocked by Cloudflare WAF Low
Bytespider (ByteDance) receives a 502 error from Cloudflare, likely a WAF block, which may unintentionally block a major AI crawler.
What to change: Review Cloudflare WAF rules to ensure Bytespider is not inadvertently blocked if access is desired.
What's working
- All major AI crawlers receive HTTP 200 — Every tested AI user-agent (GPTBot, ClaudeBot, Google-Extended, etc.) gets a full 200 response with the same content as a browser, ensuring maximum crawlability.
- Homepage delivers ~1,200 words of visible text — The homepage serves substantial text content without requiring JavaScript, making it easily indexable by AI crawlers.
- Corporate page has proper FAQPage schema — The /corporate/home/ page includes a well-structured FAQPage with three Q&A entries about how Zilch works, costs, and credit card comparison.
- US site uses FinancialService schema — The US homepage includes FinancialService structured data, correctly typing the business for search engines.
- Trustpilot reviews and NPS score displayed — The site prominently shows Trustpilot 4.6/5.0 and an NPS of 75, providing strong social proof signals for AI models.
- Arsenal F.C. partnership and major retailers listed — The site highlights a partnership with Arsenal F.C. and lists major retailers (Iceland, Morrisons, eBay), boosting brand authority.
- FCA authorisation displayed as trust signal — The site shows FCA registration number 843421, reinforcing regulatory compliance and trustworthiness.
- Sitemaps exist for UK and US regions — Both UK and US sitemaps are accessible and contain 80 and 25 URLs respectively, aiding crawler discovery.
Track zilch.com across AI search
This is one snapshot. Open the interactive report to inspect evidence, or grade another site free.