Your website has
an AI version.
We don't guess what AI thinks. We show what machines can actually find, understand and verify about your business, and what to fix before customers ask an assistant about you.
Example output. Run the check above for your own.
Google sees pages.
Machines answering questions need to understand businesses.
Your website can be online, indexed and perfectly usable by people while machines still cannot establish what you sell, who you serve, where you operate, or what you cost. We read the page a crawler is actually served and show you exactly what it can establish.
Every finding carries one of four labels
Not a disclaimer in the footer. The label is the finding.
Whether ChatGPT recommends you is UNKNOWN, and we are the checker that says that out loud. A website crawl cannot establish recommendation behaviour inside an assistant. Anyone showing you that number is guessing.
Every conclusion opens into its evidence
The report is auditable. Select any fact and see the source, the layer it came from, and the exact snippet it was read from.
Why we marked this partial. An amount was found, but no product-to-price relationship could be established.
Found in visible HTML · confidence: low
Example drawer, in the same format your report uses.
Can machines reach you?
Real requests as seven crawlers, compared against what your robots.txt promises. When the two disagree, that is the finding no robots.txt checker can give you.
Can machines understand you?
Eleven business facts read from the page a crawler is served: identity, offering, audience, location, pricing, people, trust, contact and policies. Each with its evidence.
Can machines recommend you?
Needs measurement that does not exist yet, so we do not claim it. This column ships when it can be measured, and not before.
Machine blind spots.
Important things a visitor understands about your business that machines cannot reliably establish from your site. You know these answers. Your page does not say them.
You sell accounting software specifically for freelance designers.
"Accounting software for businesses."
The gap. Your niche positioning is not explicit, so a machine answering "best accounting software for freelance designers" has no reason to name you. Worked example; your report derives yours from your own page.
The homepage says "built for modern teams" but does not specify company size, role or industry.
No clear geographic service area was found anywhere machines read.
Pricing information exists, but is not connected clearly to specific products.
Example blind spots. Your report derives yours only from measured gaps on your own page, never from a template.
Ask your website a question.
Not a chatbot. Your question is tested against the evidence on your page: what can be established, what cannot, and how to close the gap. If your site cannot support an answer, your customer does not get one either.
Four possible results: answerable, partially answerable, not answerable, or conflicting information. Every answer cites the evidence it stands on, after you run a check.
The free scan answers "what can machines understand right now?" Monitoring answers "tell me when that changes."
A deploy ships, a plugin updates, a CDN setting changes, and suddenly crawlers get a 403 or your pricing stops being machine-readable. Nothing warns you. That silence is what you pay to remove.
- Weekly automated scans and machine-understanding history
- Crawler access alerts and change detection
- Important business fact alerts and contradiction detection
- Historical AI Site View Score
Every scan already produces the versioned record this history is built from.
Managing fifty client sites?
You currently find out about a crawler block when the client does. Bulk monitoring, client workspaces, white-label reports and client-ready explanations are the agency build.
AI Site View for agenciesMake machine-readability part of every launch.
The REST endpoint this site runs on works today; a documented public version, an npm package and WordPress and Shopify integrations are the developer build, in that order.
npm install @aisiteview/site # not published yet
Free stays free.
Nothing below is charging yet. Paid tiers open when the monitoring behind them is actually running, and not a day earlier.
Free
- One-off website scan
- Crawler access, tested per bot
- Business facts with evidence
- Machine blind spots
- ASV score and score card
Monitor
Soon- Weekly automated scans
- Change and access alerts
- Machine-understanding history
- Contradiction detection
- Historical score
Pro
Soon- Multi-page and sitemap crawling
- Competitor comparisons
- Multiple sites
- Question testing at depth
- Exportable reports
Agency
Soon- Many client sites
- White-label reports
- Team access
- Scheduled reporting
- Bulk monitoring
Enterprise, with API access, large site estates and SSO, is a conversation: hello@aisiteview.com.
Recommendation visibility, competitor intelligence, AI referral tracking, public benchmarks, and a claimable AI Profile for your business. No fake dashboards and no invented percentages until each one can actually be measured.
Machines should not have to guess
what your business is.
See what your website actually tells them.