Live Case Study — Our Own Site

What Happens When You Score 95/100 — But AI Thinks You're in the Wrong City?

Our site had beautiful code, perfect bot access, and every schema type. But 3 missing tags made the AI engines think we were 4,100 km from Edmonton. Here's how we caught it, why it happens to your clients, and the 5-minute fix.

Check my site · $27

95/100
AI Visibility Score
4,100 km
Off-Target (Edmonton → Jacksonville)
3 tags
Missing Schema Properties
5 min
Time to Fix
What Happened

A Perfect Score Hiding a Fatal Flaw

TransCanadaDigital.com passed every major AI-readiness check: LocalBusiness schema, FAQ markup, GPTBot access, sitemap, fast load time. The engine returned a 95/100 — an "AI Dominant" grade.

But when the Visibility Matrix loaded competitor data, something was wrong. Instead of Edmonton agencies, the dashboard showed:

What AI Returned

  • Leapers Connect Consultancy — Jacksonville, FL 32207
  • QPS Professional Services — Unknown US location
  • Allied Consulting Firm — Unknown US location

What Should Have Appeared

  • Edmonton-based digital agencies
  • Alberta web development companies
  • Canadian competitors in the same industry
The Credibility Problem

If you send a "Local" competitor analysis to a business owner in Edmonton and it shows rivals from Florida, you lose all credibility instantly. The data is technically correct — the AI just didn't know where to look.

Root Cause Analysis

Why the AI Went to Jacksonville

The scan pipeline extracts your city from schema markup. If your JSON-LD doesn't declare a city, the API calls have nothing to "anchor" to. Without an anchor, Google Places and Serper default to the server's IP location — which, on Railway's US infrastructure, landed in St. Louis / Jacksonville.

Schema scan found Organization — but no geo

The engine extracted @type: "Organization" with an addressLocality: "Edmonton", but no GeoCoordinates were present. Without latitude/longitude, there's no mathematical proof of location.

No areaServed declared

The AI had no way to know which neighborhoods, cities, or provinces the business serves. So it searched everywhere.

No hasMap link in schema

Without a direct Google Maps URL in the structured data, the search engine couldn't "handshake" between the website and the Maps listing. The website and Google Maps were disconnected.

API defaulted to server IP geolocation

With no city anchor, the competitor lookup fell back to the physical location of the cloud server — a US-based data center. Competitors from Jacksonville and St. Louis appeared because that's where the server is, not where the business is.

"My site is beautiful, but because I was missing GeoCoordinates and areaServed, the AI literally didn't know if I was in Edmonton, St. Louis, or Jacksonville. If the machines can't pinpoint you, 'Near Me' queries will skip you every time."

— TransCanada Digital, after discovering the bug in their own audit
The 3 Invisible Gaps

Why These 3 Tags Are Non-Negotiable

Every one of these is a machine-readable signal. Without them, AI engines treat your business like a "Global Ghost" — they can see your name but have no idea where you exist.

Missing: GeoCoordinates

schema.org/GeoCoordinates — latitude & longitude

Without explicit $\text{latitude}$ and $\text{longitude}$ values in your schema, AI cannot mathematically verify your physical location. It treats you as a "floating" entity — you exist somewhere on the internet, but nowhere on the map.

What "Near Me" actually means to AI: When someone searches "web developer near me," the AI engine calculates a distance function between the searcher's coordinates and every business with declared geo data. No coordinates = you're excluded from the calculation entirely.

Impact: 100% exclusion from all proximity-based AI queries. "Near Me" searches, Google Maps AI summaries, and ChatGPT local recommendations will never include you.

Missing: areaServed

schema.org/areaServed — cities, provinces, countries

This property tells AI engines the exact geographic boundaries of your service area. Without it, the machine has to guess — and guess wrong. A plumber in Edmonton without areaServed looks identical to a plumber in Tampa to an LLM.

Why it costs you leads: When ChatGPT or Gemini recommends "the best web developer in Edmonton," it only includes businesses that have declared Edmonton as a service area. No declaration = no recommendation.

Impact: AI recommends competitors who declared their city. You get skipped — not because you're worse, but because the AI doesn't know you serve that area.

Missing: hasMap

schema.org/hasMap — Google Maps URL link

hasMap is the bridge between your website and your Google Maps listing. It's the "Social-to-Web glue" that tells search engines: "This website and this Maps pin are the same business." Without it, they're treated as two separate entities.

The disconnection effect: Your website says you're a business. Your Google Maps listing says you're a business. But without hasMap linking them, the AI sees two disconnected data points and can't build a unified "entity graph."

Impact: Google Maps ranking drops. AI can't merge your web authority with your Maps authority. Your competitors who link them get a trust multiplier you don't.
The 5-Minute Fix

3 JSON-LD Properties. 3 Minutes. Done.

This is the exact code we added to our own site's <script type="application/ld+json"> to go from "Jacksonville Ghost" to "Edmonton #1." Copy it, change the values, deploy.

Before (Broken)

"address": {
  "@type": "PostalAddress",
  "streetAddress": "Downtown Edmonton",
  "addressLocality": "Edmonton",
  "addressRegion": "AB",
  "addressCountry": "CA"
}
// No "geo" property
// No "areaServed" property
// No "hasMap" property

After (Fixed)

"address": { ... },
"geo": {
  "@type": "GeoCoordinates",
  "latitude": 53.5461,
  "longitude": -113.4937
},
"areaServed": [
  { "@type": "City",
    "name": "Edmonton" },
  { "@type": "AdministrativeArea",
    "name": "Alberta" },
  { "@type": "Country",
    "name": "Canada" }
],
"hasMap": "https://maps.google.com
  /?cid=YOUR_GOOGLE_CID"
schema.json — Complete LocalBusiness Fix 
{
  "@context": "https://schema.org",
  "@type": "LocalBusiness",
  "name": "Your Business Name",
  "address": {
    "@type": "PostalAddress",
    "streetAddress": "123 Main Street",
    "addressLocality": "Edmonton",
    "addressRegion": "AB",
    "postalCode": "T5H 3S4",
    "addressCountry": "CA"
  },
  "geo": {
    "@type": "GeoCoordinates",
    "latitude": 53.5461,
    "longitude": -113.4937
  },
  "areaServed": [
    { "@type": "City", "name": "Edmonton" },
    { "@type": "AdministrativeArea", "name": "Alberta" },
    { "@type": "Country", "name": "Canada" }
  ],
  "hasMap": "https://maps.google.com/?cid=YOUR_CID",
  "telephone": "+1-780-555-0100",
  "url": "https://yourdomain.com"
}
How to find your Google CID

Search your business on Google Maps, click your listing, and copy the URL. The cid= parameter is your unique Google Customer ID. Paste the full Maps URL as your hasMap value.

Impact Assessment

What Each Missing Tag Actually Costs You

Missing Property Status AI Consequence Severity
geo (GeoCoordinates) Missing AI cannot verify physical existence. Excluded from all proximity queries. "Near Me" = invisible. CRITICAL
areaServed Missing AI doesn't know your service area. Competitors who declared their city get recommended instead. CRITICAL
hasMap Missing Website and Google Maps are disconnected. Entity graph can't merge web + maps authority. HIGH
addressLocality Present AI knows the city name — but without geo/areaServed, it can't prove it mathematically. OK (alone: insufficient)
The Result

From "Global Ghost" to Local Domination

After adding the 3 missing tags and redeploying, the next scan anchored correctly. The AI knew exactly where to look.

Before Fix

  • Competitors from Jacksonville, FL
  • Search origin: "No city" — server IP default
  • "Near Me" queries returned 0 results
  • Maps listing disconnected from website

After Fix

  • Competitors from Edmonton, AB
  • Search origin: 🎯 Edmonton — geo-locked
  • "Near Me" queries return correct local results
  • Website + Maps unified in entity graph
The Bottom Line

You stop competing with the entire world and start dominating your local city — because the machines finally have a "Source of Truth" to trust. The fix took 5 minutes. The competitor data corrected itself on the very next scan.

For Developers & Agencies

Turn This Into a Sales Weapon

Every client you audit will have at least one of these 3 gaps. Most have all three. That's not a bug — it's your opening.

Step 1: Run Their $27 Audit

Send your prospect to transcanadadigital.com/facebook-scan. The engine will scan their schema in 60 seconds. If they're missing geo, areaServed, or hasMap, the AI Blind Spots section will flag every one in red.

Step 2: Show Them the "Wrong City" Problem

When their competitors show up from a different state or province, that's your visual proof: "See? The AI thinks you're in Jacksonville. Your customers in [their city] will never find you." This is more persuasive than any keyword report.

Step 3: Fix It in 5 Minutes, Bill for Value

The fix is 3 JSON-LD properties — you can add them in one deployment. But the value is that their business goes from invisible to recommended. That's worth a retainer, not a one-time fee.

"I built this software to find the exact invisible gaps that manual audits miss. Once the engine identifies the missing hasMap link or geo data, you can fix it in minutes and become AI-Congruent."

— Jaime Solis, Founder — TransCanada Digital

Dig Deeper

What Is AI Visibility?

The full guide to how AI search engines decide who to recommend.

 GBP Fix Guide

Fix your schema, address, CID, hours, and bot access step-by-step.

 Rich Results Explained

Stars, prices, and FAQs in Google Search — what they are and how schema triggers them.

 Answer Engine Optimization

The new SEO: how to get cited by ChatGPT, Claude, and Perplexity AI.

 Schema Doctor Pro™ Dashboard

Interactive AI intelligence dashboard with deployable fix code for your schema health.

 Plans & Pricing

From $27 audits to $19/mo full monitoring. See what fits your needs.

Is Your Site a "Global Ghost" Too?

Run the same audit we used to catch our own blind spot. It takes under 30 seconds and costs $27 — one payment, no subscription. If you're missing even one of these 3 tags, the AI engines are showing your competitors — not you.

Run a $27 Prospect Audit →
One payment. Full 5-layer scan. 30 days of re-scans. See what the machines see.