nextsense.org.au — AI Site Grade

Analysis

NextSense.org.au — AI-Visibility Audit

The site's CloudFront WAF silently strips AI crawlers of all page content — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, and Applebot-Extended all receive a 202 status with only a 2.4 KB WAF challenge page, while the browser and anthropic-ai (which bypasses CloudFront to hit nginx directly) get the full 182 KB site.

Crawler Access

The robots.txt at https://www.nextsense.org.au/robots.txt contains no AI-bot directives whatsoever — a single User-agent: * rule disallows /cpresources/, /vendor/, and /.env. This is not the problem. The problem is infrastructure: the site sits behind AWS CloudFront with an AWS WAF that issues a JavaScript challenge (status 202, ~2.4 KB shell) to most AI crawler UAs. Bytespider gets a flat 403. Only Perplexity-User and anthropic-ai reach the nginx origin server and receive full content — likely because their UAs are not yet in the WAF blocklist. The llms.txt at /llms.txt returns a 404 (full HTML page, not a text file). The sitemap index at /sitemaps-1-sitemap.xml is well-structured with 19 sub-sitemaps.

Cold-Knowledge Gap

The LLM knows NextSense as a 2020 merger of RIDBC and Australian Hearing Services that faced 2023 restructuring criticism and staff redundancies. The site itself never mentions the merger, the RIDBC legacy name (except buried in the history timeline), or any restructuring. The homepage and all key pages present a unified "NextSense" brand with no acknowledgment of the 2020 merger or the former RIDBC identity. The cold knowledge also references "custom hearing aids" as a notable product — the site does not prominently feature hearing aid sales, focusing instead on cochlear implants, early intervention, and education.

Schema Posture

JSON-LD is present on every page but has a structural flaw: the @graph array includes a LocalBusiness entry for the marketing agency Simple Integrated Marketing (https://simple.com.au#creator) as a publisher and creator of every page. This agency schema (with full address, hours, and pricing) is injected into the page-level schema, diluting the brand's own identity. The homepage uses WebSite schema; service pages use WebPage; the vision early-intervention page uses MedicalWebPage. No FAQPage schema exists on the CI FAQ page — that page is essentially a shell containing only a link to an Adobe Acrobat PDF hosted off-domain. No Article or NewsArticle schema on the news listing or individual story pages.

Content Signals

The homepage is content-rich (1,144 words) with clear H1/H2 structure and strong answer-format signals: lists of services, numbered statistics ("400+ people receive cochlear implants", "5,200+ children and adults"), and a prominent "Redefine what's possible" tagline. However, the CI FAQ page (/services/cochlear-implant-program/ci-faq) contains only 14 words of visible text — it is a landing page that redirects users to an external Adobe Acrobat PDF for the actual FAQ content. This means the site's most FAQ-heavy page is invisible to AI crawlers even if they bypass the WAF. No comparison tables, no structured FAQ markup, no inline Q&A.

External Signals

The site has no discoverable external press coverage, Reddit threads, or review pages indexed in search results for the NextSense brand. The DNS records show integrations with Microsoft 365 (Outlook mail), Zoom, Atlassian, Adobe, Intacct, Nitro, Autopilot, SendGrid, Mailchimp, and Sophos — a complex tech stack. Social profiles (Facebook, LinkedIn, Twitter/X, Instagram) are linked but no social proof or review aggregator citations appear on the site. The sameAs schema references only three social URLs (missing LinkedIn, which is linked in the page footer).

Findings

1. CloudFront WAF blocks most AI crawlers with JavaScript challenge High

   GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, and Applebot-Extended all receive a 202 status with a 2.4 KB WAF challenge page instead of full content. Only Perplexity-User and anthropic-ai bypass the WAF and reach the origin server.

   What to change: Update the AWS WAF rules to allow AI crawler user agents (GPTBot, ClaudeBot, etc.) to bypass the JavaScript challenge and serve full HTML content.

2. CI FAQ page contains only 14 words and links to external PDF High

   The cochlear implant FAQ page at /services/cochlear-implant-program/ci-faq has only 14 words of visible text and redirects users to an external Adobe Acrobat PDF for the actual FAQ content. This makes the page invisible to AI crawlers even if they bypass the WAF.

   What to change: Replace the shell page with inline FAQ content using structured FAQPage schema, or at minimum include the full text on the page.

3. JSON-LD includes marketing agency as publisher and creator Medium

   Every page's JSON-LD @graph array contains a LocalBusiness entry for Simple Integrated Marketing (simple.com.au#creator) as publisher and creator, with full address, hours, and pricing. This dilutes NextSense's own brand identity in structured data.

   What to change: Remove the agency's LocalBusiness entry from page-level schema or move it to a separate, non-page-level context.

4. No FAQPage schema on CI FAQ page Medium

   The CI FAQ page lacks FAQPage structured data, which would help AI assistants extract and present Q&A content directly.

   What to change: Add FAQPage schema with the questions and answers from the PDF.

5. No Article or NewsArticle schema on news pages Medium

   The news listing and individual story pages do not use Article or NewsArticle schema, reducing their chance of appearing in AI-generated news summaries.

   What to change: Add Article or NewsArticle schema to news and story pages.

6. llms.txt returns 404 Low

   The /llms.txt endpoint returns a 404 HTML page instead of a plain text file, missing an opportunity to guide AI crawlers to key content.

   What to change: Create an llms.txt file listing key pages for AI crawlers.

7. robots.txt lacks AI-bot directives Low

   The robots.txt file has no specific rules for AI crawlers, only a generic User-agent: * rule disallowing a few paths. This is not harmful but misses the chance to explicitly allow AI bots.

   What to change: Add explicit allow directives for AI crawlers (e.g., GPTBot, ClaudeBot) to ensure they are not blocked by future WAF changes.

8. Site does not mention 2020 merger or RIDBC legacy Medium

   The LLM cold knowledge identifies NextSense as a 2020 merger of RIDBC and Australian Hearing Services, but the site presents a unified brand with no mention of the merger or RIDBC except in the history timeline. This creates a disconnect for AI assistants that may reference the legacy names.

   What to change: Add a brief mention of the merger and legacy names on the homepage or about page to align with external knowledge.

9. No discoverable external press coverage or reviews Medium

   Web searches for NextSense Australia returned zero results for press, Reddit threads, or review pages. This limits external signals that AI models use to validate authority.

   What to change: Encourage media coverage, customer reviews, and community discussions to build external signals.

10. sameAs schema missing LinkedIn URL Low

    The JSON-LD sameAs property includes only three social URLs (Facebook, Twitter, Instagram) but omits LinkedIn, which is linked in the page footer.

    What to change: Add the LinkedIn URL to the sameAs array in the JSON-LD.

What's working

• Homepage is content-rich with clear structure — The homepage has 1,144 words, clear H1/H2 headings, and strong answer-format signals like lists and statistics, which help AI assistants extract key information.
• Sitemap index is well-structured with 19 sub-sitemaps — The sitemap index at /sitemaps-1-sitemap.xml is properly formatted and contains 19 sub-sitemaps, aiding crawler discovery.
• JSON-LD structured data present on every page — All pages include JSON-LD with WebSite, WebPage, or MedicalWebPage schema, providing a baseline of structured data.
• anthropic-ai crawler bypasses WAF and gets full content — The anthropic-ai user agent reaches the nginx origin server directly and receives the full page content, ensuring Claude can index the site.
• Social media profiles linked in footer — The site links to Facebook, LinkedIn, Twitter/X, and Instagram profiles, providing social signals.