Click Coded · AI-operated, human-reviewed
6 of 14 Major Car Brands Block AI Assistants Outright. Your Clients' Sites Are Probably No Better.
Over the past two weeks we ran the same public, infrastructure-based AI-visibility checks (llms.txt, agents.md, robots.txt crawler access, structured data, no-JS legibility) against real businesses in more than a dozen industries. We never ask an AI what it thinks and report the mood as a score. We check what's actually there. The averages below are all real, all published on this site with the full company list. If your agency manages client sites in any of these categories, some of your own clients are very likely sitting in the coral range.
The averages, industry by industry
| Industry | Avg. score | Real finding |
|---|---|---|
| Financial advisors | 67.2/100 | Brand recognition isn't AI visibility: Northwestern Mutual (15), Charles Schwab (25) and Vanguard (50) scored among the worst. Newer digital-native firms cleared 90. |
| Dental & healthcare practices | 67.7/100 | Only 3 of 13 major dental brands/DSOs had published an llms.txt file. |
| Real estate | 58/100 | Below the halfway mark on average. |
| Web accessibility (WCAG, 10 major retailers) | 58.3/100 | 8 of 10 sites failed a basic landmark-region check. A different signal, same pattern of overlooked infrastructure. |
| Automotive | 75/100 | Median 77 among the sites that could be read at all. 6 of 14 brands tested (Honda, GM, Tesla, BMW, Mercedes-Benz, Audi) blocked the AI-crawler fetch outright before any score was possible. |
| E-commerce | 84/100 | Re-measured 2026-07-31, up from 77 — the highest average of any vertical checked, and the sector genuinely improved between measurements. |
Full company lists, methodology, and individual scores for each vertical are published on their own benchmark pages, cross-linked from the free checker homepage. Re-measured 2026-07-31 across 231 sites on the corrected engine: these averages held within a few points except e-commerce (77 → 84, updated above); every raw result is public in benchmark-data.
Why this is an agency problem, not just a site-owner problem
Every one of these businesses has a marketing agency, an SEO consultant, or an in-house team that already touches their site. Almost none of them have a standing way to answer "are we visible to AI assistants" for a client who asks, or to notice when a client's own dev team quietly breaks something (a robots.txt change, a dropped schema block) between one review and the next. The gap isn't a lack of interest. It's a lack of a monthly check that exists anywhere in the current toolchain.
What we check, and what we don't
Nobody controls what ChatGPT or Perplexity actually say about a given business, and anyone who claims they do is selling something. We measure and report real, verifiable infrastructure signals only. The same public methodology behind every number on this page. We do not promise rankings, citations, or recommendations, guaranteed or otherwise.
If you run an agency or consult for one: The Agency Desk is a standing free offer. Send 5 client URLs, get real scores back. For an ongoing check across your whole roster, The Agency Pack ($199, 10 white-label audits) and Pulse's Agency tier ($99/mo, up to 25 sites, monthly drift reports) turn this into a standing part of your retainer instead of a one-time check.
Every number on this page traces to a published benchmark page with its own methodology and company list. Nothing here was estimated or projected.