Real Results
What AI Actually Says About
Regional Australian Businesses
Not scores on a dashboard. Verbatim AI responses — before and after — for real regional Australian businesses.
68
businesses audited
3
AI platforms tested per audit
2 of 6
avg. criteria failing before audit
48 hrs
from URL to delivered report
Most Businesses Are Invisible to AI — and Don't Know It
When someone asks ChatGPT “find me a real estate agent near me” or Perplexity “who does property management in regional Victoria,” the AI doesn't search Google. It draws on structured content it has already indexed — and for most regional SMBs, that content is incomplete, ambiguous, or missing entirely.
89%
of SMB websites are inconsistently represented across ChatGPT, Perplexity, and Google AI — meaning each platform tells a different story about the same business.
Every vague AI answer sends a buyer somewhere else.
When AI hedges — “likely,” “may require” — buyers choose the competitor who got a clear answer instead. There's no middle ground.
The businesses below are typical: established, reputable, with real clients — but AI tools were giving buyers incomplete or inaccurate answers about them every day. Here's what that looked like, and what changed.
Case Study #001
Regional Real Estate Agency
The Problem
This agency has served their regional Victorian community for decades. Their website clearly lists services, locations, and contact details. But when we ran AI simulations before the audit, this is what buyers were getting:
ChatGPT response — before audit
“Local real estate presence. Likely multi-service offering. Contact details may require extra navigation.”
Likely. A business that has operated in the region for decades — and the AI was guessing about what they do.
What the Audit Found
The website itself was fine. The problem was structural — two things that AI platforms specifically look for were absent:
Missing buyer-intent language
AI platforms build answers from FAQ-style, question-and-answer content. The site had service descriptions but no content structured around what buyers actually ask.
No trust constraints
Without explicit guidance, AI tools invent caveats — "contact details may require extra navigation." A trust constraints block tells AI exactly what it can and cannot claim.
The Fix
We built a custom llms.txt file — a structured document that tells AI platforms exactly who this business is, what they do, where they operate, and how to answer buyer questions with confidence.
- A high-intent query map — 8 buyer questions with structured, accurate answers
- Service architecture — residential sales, rentals, rural/livestock clearly segmented
- Location coverage — primary town and regional service area made explicit
- Trust constraints — what AI can and cannot claim on behalf of the business
- Contact pathway — canonical URL, not just a phone number
- Entity/synonym map — business name variations and local shorthands all mapped
What AI Says Now
Responses simulated using 8 buyer-intent queries across ChatGPT, Perplexity, and Google AI ecosystems.
“Summarise this business for a buyer”
Before
“Local real estate presence. Likely multi-service offering. Contact details may require extra navigation.”
After
“Regional Victorian real estate agency covering residential sales, rentals, and rural/livestock property. Clear contact-first path for buyer enquiries via the website.”
"Likely" — the AI was guessing about a business that has operated in the region for decades. After the audit, every response is confident and specific.
“What areas do they service?”
Before
“One town name only.”
After
“Primary town and broader regional coverage — including rural and livestock property across the surrounding area.”
One word vs a full service area. Before the audit, anyone asking AI about regional coverage got an incomplete answer.
“What should I know before contacting them?”
Before
“No specific information available. Standard real estate agency terms likely apply.”
After
“Specialises in regional residential, rental, and rural/livestock property. Confirm service area and property type match before enquiring. Contact via the website for the fastest response.”
Before: AI deflected with a generic disclaimer. After: it gave a buyer the exact context they needed to make a confident enquiry.
“How do I contact them right now?”
Before
“Call [phone number].”
After
“Visit the website contact page for the fastest response — phone and enquiry form both available.”
A phone number alone is a dead end for anyone browsing on their phone at 9pm. The after response gives a web-first path that works any time.
AI Readiness Scorecard
Bottom Line
A 4/6 score means AI tools have partial information — enough to find the business, not enough to confidently recommend it. Every buyer query that gets a hedged or incomplete answer is a potential referral lost to a competitor. A 6/6 score means AI returns complete, confident, buyer-ready answers every time.
What Is AI Saying About Your Business?
Start with a free pre-audit assessment. We'll tell you where the gaps are before you spend anything.