pointsbet.com.au — AI Site Grade

Analysis

I have all the data I need. Here's the audit.

---

PointsBet.com.au — AI-Visibility Audit

Every page on pointsbet.com.au returns zero visible text to any HTTP client, including every major AI crawler. The entire site is a single-page application (SPA) shell — a <div id="splash"> with a spinner SVG and a single <script> bundle (core.YP5RHC2P.js). No server-side rendering, no prerendered fallback, no static content. All 11 AI bots tested (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User, anthropic-ai, Bytespider, Applebot-Extended, Perplexity-User) receive the same 9,428-byte HTML shell with 0 words of readable content. The site is not blocking bots — it is structurally invisible to them.

Crawler Access

The robots.txt contains no AI-bot-specific directives. A single User-agent: * rule disallows /account/, /deposit/, /withdraw/, /bonuses/, /rewards/, /invite/, and URL parameters containing promo or utm_. All AI crawlers are permitted to crawl the entire site — but every page they reach, from /about-us to /privacy to /how-inplay-betting-works, returns the identical empty SPA shell. The llms.txt endpoint returns the same SPA HTML (served as text/html, not text/plain), meaning it is not a real llms.txt file — it is a 404-equivalent served by the SPA router. The site runs on Cloudflare (A records at 104.16.216.51/217.51) with HSTS, X-Frame-Options SAMEORIGIN, and a 600-second cache TTL.

Cold-Knowledge Gap

A frontier LLM queried cold about PointsBet Australia knows: the brand was founded in Australia in 2015, offers a "PointsBetting" feature where payouts scale with margin of victory, expanded into the US via NBC Sports partnerships, was fined $30,000 by NSW regulators in 2022 for accepting bets from a self-excluded person, and was acquired by BlueBet in 2023. None of this information exists on the live site. The site's meta description ("Australia's fast growing online bookmaker") and OG tags are the only textual signals AI engines can extract. The LLM's prior knowledge — including the BlueBet acquisition, the regulatory fine, and the US expansion — comes entirely from off-domain sources. The site itself provides zero structured data, zero JSON-LD schema, zero FAQ markup, zero comparison tables, and zero heading structure on any page.

Schema and Content Posture

No page on the domain contains any JSON-LD schema of any type. No Organization, SportsOrganization, WebSite, FAQPage, Product, or BreadcrumbList schema is present. The homepage has no <h1>, no <h2>, no internal links extractable from the static HTML. The sitemap lists 203 URLs including deep sport-specific pages (/sports/aussie-rules/AFL, /sports/basketball/NBA), but every one of those URLs resolves to the same empty shell. The Wayback Machine has captured 560 snapshots of the homepage since 2017 — all of them the same SPA shell. The site has never served server-rendered content to archival crawlers.

External Signals

PointsBet Australia has significant off-domain footprint — the BlueBet merger, NBC Sports US partnership, AFL/NRL sponsorship deals, and regulatory fines are all documented externally. However, the site's own domain has zero discoverable external mentions via search, and no Reddit threads, review sites, or press articles surfaced in testing. The brand's external reputation is carried entirely by third-party sources that AI models have ingested during training; the domain contributes nothing to that narrative.

Findings

1. Entire site is an empty JavaScript shell with zero visible text High

   Every page on pointsbet.com.au returns a 9,428-byte HTML shell containing only a splash div with a spinner SVG and a single script bundle. No server-side rendering, no prerendered fallback, and no static content exists. All 11 AI bots tested receive zero words of readable content.

   What to change: Implement server-side rendering (SSR) or static prerendering for all pages so that crawlers receive meaningful HTML content. Alternatively, use dynamic rendering to serve pre-rendered snapshots to known bot user agents.

2. Zero JSON-LD schema of any type on any page High

   No page on the domain contains any JSON-LD schema markup. Missing Organization, SportsOrganization, WebSite, FAQPage, Product, and BreadcrumbList schemas. This prevents AI engines from extracting structured information about the brand, its offerings, or its content.

   What to change: Add JSON-LD structured data for Organization, WebSite, SportsOrganization, and relevant page types (e.g., FAQPage for help content, Product for betting offers).

3. No heading tags (h1, h2) on any page Medium

   The homepage and all other pages lack any heading elements in the static HTML. This eliminates a key signal for AI crawlers to understand page structure and topic hierarchy.

   What to change: Add descriptive h1 and h2 tags to each page, reflecting the page's primary topic and sections.

4. llms.txt endpoint serves SPA HTML instead of plain text Medium

   The /llms.txt URL returns the same SPA HTML shell (content-type text/html) rather than a plain text file. This means the site does not have a functional llms.txt file, missing an opportunity to provide AI crawlers with a curated summary of available content.

   What to change: Create a proper llms.txt file at /llms.txt that lists key pages and a brief site description, served as text/plain.

5. Meta description is generic and lacks substantive content Medium

   The homepage meta description reads 'Australia's fast growing online bookmaker' with no additional context about offerings, features, or value proposition. This is the only textual signal AI engines can extract from the page.

   What to change: Expand the meta description to include key differentiators (e.g., PointsBetting, sports coverage, promotions) and ensure it is unique per page.

6. No extractable internal links in static HTML High

   The static HTML of the homepage contains no anchor tags or internal links. All navigation is rendered client-side via JavaScript, making it invisible to crawlers that do not execute JS.

   What to change: Include static HTML links (e.g., in a sitemap or footer) that point to key pages, or implement SSR to render navigation links in the initial HTML.

7. No FAQ, comparison tables, or structured content on any page Medium

   The site lacks any FAQ sections, comparison tables, or other structured content formats that AI engines commonly extract for featured snippets and knowledge panels.

   What to change: Create FAQ pages with JSON-LD FAQPage schema, and add comparison tables for betting options or promotions.

8. Zero discoverable external mentions of the domain via search Medium

   Multiple web searches for the domain and brand returned no results, indicating the site has minimal to no backlink profile or external citations that AI crawlers can use to validate authority.

   What to change: Develop a link-building strategy to earn mentions from reputable sports and betting sites, and ensure the site's content is indexable so that external links can be discovered.

9. No Reddit threads or review site mentions found Low

   Searches for PointsBet Australia on Reddit and review sites returned no results, indicating a lack of community discussion and third-party validation that AI models often use as signals.

   What to change: Encourage user reviews on platforms like Trustpilot and engage in relevant Reddit communities to build organic mentions.

10. OG tags are minimal and do not include rich media Low

    The homepage includes basic Open Graph tags (title, description, image) but no additional tags like og:video, og:audio, or article-specific tags that could enhance social and AI previews.

    What to change: Expand OG tags to include richer metadata such as og:type, og:site_name, and page-specific images.

11. All 203 sitemap URLs resolve to empty SPA shell High

    The sitemap lists 203 URLs including deep sport-specific pages, but every one returns the same empty JavaScript shell with no content. This means the sitemap does not help crawlers discover meaningful content.

    What to change: Ensure that all sitemap-listed URLs serve unique, meaningful HTML content via SSR or prerendering.

12. All Wayback Machine snapshots show empty SPA shell since 2017 Medium

    The Wayback Machine has captured 560 snapshots of the homepage since 2017, all showing the same empty SPA shell. The site has never served server-rendered content to archival crawlers.

    What to change: Implement SSR or dynamic rendering so that archival crawlers can capture meaningful content.

What's working

• Robots.txt allows all AI crawlers to access the entire site — The robots.txt file contains no AI-bot-specific disallow rules. All major AI crawlers are permitted to crawl the entire domain, which is a positive foundation for AI visibility once content is made accessible.
• Site uses Cloudflare CDN with strong security headers — The site is served via Cloudflare with HSTS, X-Frame-Options SAMEORIGIN, and a 600-second cache TTL, providing good performance and security posture.
• Sitemap exists with 203 URLs including deep sport-specific pages — The sitemap is accessible and lists 203 URLs, including detailed sport-specific pages (e.g., /sports/aussie-rules/AFL). This provides a roadmap for crawlers once content is rendered.
• LLM cold knowledge contains substantial brand details — A frontier LLM queried cold about PointsBet Australia knows key facts: founded in Australia in 2015, PointsBetting feature, US expansion via NBC Sports, regulatory fine in 2022, and acquisition by BlueBet in 2023. This off-domain knowledge can be leveraged once the site provides its own content.
• No AI bots are blocked by robots.txt or server configuration — All 11 tested AI bots received a 200 HTTP status and were not blocked by any access control. The site's invisibility is purely due to lack of content, not access restrictions.
• Wayback Machine has consistent archival history since 2017 — The Wayback Machine has captured 560 snapshots of the homepage since 2017, indicating the site has been consistently accessible to archival crawlers, even though the content is empty.
• Basic Open Graph tags are present on the homepage — The homepage includes og:title, og:description, and og:image tags, providing minimal metadata for social sharing and AI previews.
• DNS and hosting infrastructure is stable with multiple records — The domain has two A records, two NS records, and seven TXT records, indicating a well-configured DNS setup with Cloudflare hosting.