How to Tell If Anyone's AI Assistant Is Actually Citing You
The only test that counts is whether the assistants name you when an Australian buyer asks. Here is the monthly check, the CRM question that ties it to pipeline, and how much to trust the scores.
Tool 3: The One Everyone Forgets, Your Own CRM
Here's the lowest-tech, most underused AEO measurement tool available, and you're already paying for it: a question on your enquiry form.
Add "How did you first hear about us?" to your forms, with "AI tool (ChatGPT, Perplexity, etc.)" as one of the options. That answer becomes a contact property in HubSpot. Which means you can report on it, trend it over time, and watch AI-sourced enquiries grow (or not) as an actual line on a chart rather than a hunch.
This is the measurement that matters most, because it's the only one tied directly to pipeline. The Grader tells you what the models think. The CRM tells you whether that's turning into actual clients contacting your actual business. One is interesting. The other pays the bills.
Allow for the lag. There's usually a 30-to-60-day gap between AI visibility improving and enquiries showing up, because that's the normal research-to-inquiry delay. So correlate your AEO score improvements with lead volume a month or two later. Same-week comparisons will tell you nothing.
The Check You Run By Hand
The tooling in this category is young. There is no citation tracker anyone should bet a budget on yet, and the Australian market is the last place any of them get right. So the most reliable test is still the one you run yourself: ask the questions your buyers ask, in the assistants your buyers use, and write down who gets named.
Pick five questions in your buyer's words. Ours are things like "who are the best HubSpot partners in Brisbane" and "what does RevOps cost in Australia". Run each one through Claude, ChatGPT, Perplexity and Gemini once a month, in a fresh chat with no history, and log two things: whether Neighbourhood came back, and which competitors did. That's 20 prompts and about ten minutes.
Australian questions behave differently from the generic ones the global tools are tuned on, which is why Australian brands specifically go missing in AI search. An assistant that names you for "HubSpot partner" and drops you for "HubSpot partner Brisbane" is telling you something no score out of 100 will.
Worth knowing what market you're doing this in: ABS figures put AI use among Australian businesses at 12%. Most of your competitors are not watching this yet, and the buyers who are already building a shortlist in an assistant are ahead of the industry they're shopping in.
Putting It Together: A Simple Measurement Cadence
You don't need a dashboard with forty metrics. You need a cadence you'll actually keep. Here's a sensible one for a 60 to 200 person business:
Monthly: Run the free AEO Grader on yourself and your top two competitors. Log the five dimension scores in a simple sheet. You're watching the trend line, not the individual number.
Monthly: Run your five buyer questions through the four assistants by hand and log who got named.
Monthly: Check your "How did you hear about us?" contact property in HubSpot. Note how many enquiries cited an AI tool. Watch the direction over time.
Ran the grader and unsure what the score means?
A number without a baseline reads the same every month.
Quarterly: Do a proper review. Has your share of voice moved against competitors? Are AI-sourced enquiries trending up? Which content did you publish or refresh, and did anything correlate with a score change 30 to 60 days later?
If you're on the paid tool: Add a weekly glance at citation sources, and treat each "competitor cited instead of you" as a potential content brief.
That's it. Half an hour a month and a proper look each quarter will tell you more than most businesses know about their AI visibility. If none of the groundwork is done yet, measuring is premature: start with the foundational AEO playbook this piece builds on and come back to the scores in a quarter.
What "Good" Actually Looks Like
People want a target number. The honest answer is that the number matters less than the direction and the context.
A score of 45 that's climbing steadily, in a category where your competitors sit at 30, is a good position. A score of 70 that's been flat for six months while a competitor climbed from 40 to 65 is a worse position, even though the raw number is higher. AEO is relative and it's directional. Read it that way.
It's also a different job from SEO, which changes what's worth measuring in the first place. How AEO differs from SEO and GEO is the piece to read before you set targets against the wrong metric.
The clearest sign that it's all working is a new enquiry that says "ChatGPT recommended you", and then it happens again the next month, and the month after. That's the result the scores are a proxy for. When the proxy and the pipeline both point the same way, you're winning.
So Go Check
Take five minutes. Run the AEO Grader on your brand and your closest competitor. Look at the gap.
Then go add that one question to your enquiry form, because in six months you'll want the data and the only way to have it then is to start collecting it now.
The businesses that win AI search measured honestly, kept doing what moved the number, and stopped doing what didn't.
Want help making sense of your AEO numbers, or actually moving them? Message us. We'll show you where you stand and what's worth doing about it, and it sits inside our websites and experience work.
Catch our latest HubSpot masterclasses and tutorial shorts on YouTube, or join us over on Facebook for a healthy dose of marketing reality. No fluff, just the good stuff.
Can you tell if AI search is sending you anyone?
Most CRMs record the visit and lose where it came from.
Happy HubSpotting.