GeoFoundry
§ sample report

This is what you actually get

No mystery deliverable. This is the assay, section by section, with the kind of findings each one contains. Every number in it traces back to something we can put in front of you.

note on the samples
# examples below are illustrative and labeled as samples.
# your report is built from real, recorded runs on your business.
§ report structure

Nine sections.
Every one with its evidence.

Each section below appears in every GeoScan, in this order. No section makes a claim it can't show the reading for.

01

Executive Summary

The strategic story in one page — where the business stands with AI, what's costing it visibility, and what changes that.

# sample

“The clinic is effectively invisible to AI assistants. Across sampled runs, competitors were recommended in the majority of relevant queries while the clinic appeared in none. The cause is not reputation — it's readability: the website gives AI nothing to extract, cite, or trust. The fixes are concrete and inside the clinic's control.”

02

GEO Scorecard

Every factor score with its value, weight, and a plain-language reason. Visibility on one side, extraction readiness on the other.

# sample

Sample factors: AI Mention Rate · Citation Rate · Brand Accuracy · Competitor Comparison · Answerability · Structural Semantics · Machine Readability · Topical Coverage · Technical SEO Health

03

Site-Level Analysis

Factor by factor: what we detected on the site, what's missing, and the specific issue each section closes with.

# sample issue

“No structured data exists on any service page. AI systems must guess what the business does — and in sampled runs, they guessed wrong.”

04

AI Query Simulation

Real customer questions run against the AI systems, classified Answerable / Partially Answerable / Not Answerable from the live site — each with the recorded response and the gap.

# sample

“Q: Do they offer emergency service? → Not Answerable. The site mentions ‘emergencies’ once, in a footer. No hours, no process, no page. AI cannot answer what the site does not state.”

05

Strengths & Weaknesses

What's genuinely working — and what's actively costing visibility.

# sample

Sample strength: strong review profile AI can see. Sample weakness: services buried in PDFs AI cannot read.

06

Priority Recommendations

Tiered Critical / High / Structural. Each with the specific page, the specific fix, why it matters, and expected impact.

# sample (critical)

“Add LocalBusiness + Service schema to the five service pages. Why: AI currently cannot confirm what you do or where. Impact: directly feeds entity accuracy and citation eligibility.”

07

Strategic Insight

The narrative synthesis — what the findings mean together, and the order of operations that compounds.

# sample

“Fix readability before chasing mentions. Visibility built on a site AI can't read is rented, not owned.”

08

Final Verdict

The bottom line, stated plainly.

# sample

“Today, AI recommends your competitors because they are easier to read — not because they are better. That is fixable.”

09

References & Evidence Trail

What was crawled, what was detected, what was absent, and every saved AI response behind every score — word for word.

# sample

42 pages crawled · sampled prompt runs recorded across 2 AI systems · sampling configuration disclosed · raw responses included in full.

§ why it's built this way

Gauges, not vibes.
A report you can verify.

  • Every claim traces to evidence: a crawled page, a detected or absent signal, or a recorded AI run. Nothing is written as plausible fiction.
  • Recommendations name specific pages, specific schema types, specific content — implementable by your team, ours, or your existing provider.
  • The sampling configuration is disclosed in the report itself: which AI systems, how many prompts, how many runs.
Get your own report What's included & pricing
evidence trail · every claim
— every score resolves to one of these
crawled_page      url + timestamp
detected_signal   what was found, where
absent_signal     what was looked for, missing
recorded_run      prompt + verbatim response

# sampling config printed in the report:
# which systems · how many prompts · how many runs
§ free check

Find out what AI says about your business.

Free one-page AI Visibility Check. No call required — we run the scan and email you the result.

Every score backed by saved AI responses

No obligation. We only use your information to deliver your report.