Skip to main content
Fieldnote

Technical GEO audit for marketers

Can Claude read your website?

See who gets through, what they receive and what to fix.

Bot accessSSR vs CSRPage typesSchemaTrust evidence

Get an invite within 48 hours. We will not crawl your site until you confirm.

Public websites only. No crawl starts yet. Read the data notice.

Example access check
Interactive
Blocked

OpenAI cannot read this page

Response

HTTP 403

Content received

0%

The security layer stopped the request.

Next: Allow the crawler through the WAF.

Example Fieldnote test request. It does not claim access from a vendor's private network.

SSR vs CSR

What visitors see is not always what AI receives.

Client-side rendering waits for the browser to build the page. Server-side rendering sends the important content immediately. Fieldnote compares both.

Example rendering check

example.com/pricing

Compare rendering methods
Visitor's browser100% visible

Plans

Choose the plan that fits

Starter

$29

Growth

$79

Scale

$149

11 pricing questions answered below the plans
Direct crawler request12% received

Plan content is not in the first response

Key content received

12%

Plans 0 of 3

Prices 0 of 3

FAQs 0 of 11

Important content arrives too late

The browser builds the pricing page after JavaScript runs. A direct crawler request receives the navigation and footer, but none of the plan details.

First action: Send plan names, prices and FAQs in the first response.

Example comparison from a Fieldnote request, not a vendor-owned bot network.

Smarter sampling

Audit the few pages that represent the whole site.

Fieldnote groups pages by purpose, selects representative examples and applies the right checks to each template.

1,284

URLs discovered

6

Page types

12

Pages selected

Choose an example page type
96% classification confidence

Insight

/insights/material-guide

Found

412 URLs

Audited

3 pages

Confidence shows how certain Fieldnote is about the page type. It is not a page-quality score.

Why this page was selected

The site’s main editorial template and content structure.

Checks applied to this template

Author proof
Sources
Updated date
Article markup

Example finding

The author is named, but relevant experience is not shown.

Schema and trust

Show AI facts it can verify.

Fieldnote checks the labels in your code and the visible proof behind your claims, then explains what is missing.

Choose an evidence check

Readiness

61 / 100

Schema

Code labels facts such as product name, price and availability so systems do not have to guess.

Product name and SKUFound
Price and currencyMissing
AvailabilityMissing
BreadcrumbsFound
Brand connectionPartial

So what

AI can identify the product, but may not confidently report what it costs or whether it is available.

First action

Add price, currency and availability to the product markup.

Readiness summarises evidence found on the page. It is not a ranking or AI visibility guarantee.

Why it matters

The first HTML response shapes what AI answers.

The goal is not just crawl access. It is about prompts that convert into qualified opportunities.

What this means for marketing outcomes

AI sees what is in your first HTML response, not what only loads in a browser.

Prompt

What plan fits a 25 person team and why is this offer different?

Rendering in crawl test

Client-rendered (CSR)

Why this changes outcomes

Sales loses qualified leads to unclear recommendations.

AI-visible result

What crawler sees

Plan names and prices are not present in the first crawl response.

11% of key facts are present in first response.

Example bot answer

A bot can only give a partial comparison and suggests a broad offer without the right plan.

Better first-response facts usually raise confidence and improve lead quality.

Clear scores. Clear next steps.

Know the score. Know the fix.

Click a score to see the finding and first action.

GEO readiness

64 / 100

Needs work
38 / 100

OpenAI blocked

Security returned HTTP 403.

First action

Allow the crawler at the WAF.

Private beta

Find out what AI search can actually read.

Join the waitlist
Data notice

The waitlist stores your work email, the public website origin you submit, the signup source, submission timestamps and a repeat-submission count in a private Cloudflare D1 database. There is no public signup list. The data is used only to review beta requests and contact you about access; it is not published or sold. Records are kept while the beta is being developed or until you ask for removal. Joining does not start a crawl or store crawler response bodies. If you later join an audit, limited public-page evidence may be sent to OpenAI for E-E-A-T review with response storage disabled. Do not submit private or confidential URLs. For access, deletion, or data questions, email support@wthisgeo.com.