You can do everything right and still have no idea whether it worked. There is no console for assistants, no impressions report and no ranking chart. So AI search visibility becomes the first thing in marketing you cannot simply look up, which leaves owners choosing between paying for a tool they cannot evaluate and ignoring the question entirely. There is a middle option, and it costs nothing but an hour a quarter.
Why the usual reports do not help
Analytics measures visits. If someone asks an assistant who to call in Medford, hears your name and dials the number, no visit occurs. The most valuable outcome produces the least evidence.
Referral data is patchy for the same reason. Some assistants pass a recognisable source when a link is clicked, many interactions never involve a click at all, and the ones that matter most for a service business often end in a phone call.
So the honest position is that this cannot be measured precisely today. What it can be is tracked directionally, which is enough to make decisions with.
The manual audit that actually works
Build a list of ten to fifteen questions a real customer would type. Not keywords, questions. Who does this work near Ashland. What does this cost in southern Oregon. How do I choose between these two options. Is this a job I can do myself.
Run them through the assistants your customers plausibly use, and record four things for each: whether you were named, which competitors were named, which sources were cited, and whether anything said about you was wrong.
Repeat the identical list quarterly. The value is entirely in the comparison, which is why the wording must stay fixed and why a single run tells you very little.
Reading the results honestly
Three patterns are worth acting on.
Nobody local is named. Common in smaller markets, and the best news on this list. The category is open, and consistency established now compounds before competitors bother.
Competitors are named and you are not. They are described more consistently across the web than you are. That is a corroboration problem rather than a content problem, and our guide on how to fix NAP issues and clean up local citations is the first place to look.
You are named but the details are wrong. The most urgent case, because a confidently stated wrong phone number does active damage. Trace it to the outdated source and correct it there.
AI search visibility: the proxy metrics that do exist
While direct measurement is unavailable, several numbers you already have move in related ways.
- Calls and direction requests from your profile. These capture the outcome even when no website visit happens. If you are not tracking calls by source, our guide to call tracking for local businesses across Maps, ads and organic covers setting it up.
- Branded search volume. If more people search your business name directly, something is putting it in front of them. This is one of the better indirect signals available.
- Impressions against clicks. Impressions holding while clicks fall usually means answers are being displayed rather than visibility being lost. Our piece on using Google Analytics data to improve SEO performance covers reading that pattern.
- What new enquiries say. Ask how people found you. It is unfashionable and it is currently better data than any dashboard.
None of these prove anything on their own. Read together, over a quarter, they are enough to tell you whether the work is landing.
Tools that sample assistant answers repeatedly and report how often you appear do exist, and they are legitimate. Two caveats matter before you spend.
They report frequency of appearance across sampled prompts, not a ranking position. Anyone presenting it as a rank is misrepresenting how these systems work. And their value depends entirely on whether the prompts they sample resemble what your customers actually ask, which for a regional service business is often not the case.
For most local businesses in the Rogue Valley, the manual quarterly audit gives you the same decision-making information for free. Revisit the question when you have exhausted the fundamentals.
What to do with what you learn
The audit is only useful if it changes something. If you are absent, the answer is almost always consistency and corroboration rather than more content. If details are wrong, fix the source. If competitors dominate, look at where they are mentioned that you are not.
Keep the fundamentals current underneath all of it, since a complete profile remains the most reliable description of your business anywhere. Our GBP optimisation checklist for service area businesses covers that, and our guide to conversion rate optimisation covers making sure the enquiries you do get convert.
If you would rather have the audit run properly and turned into a list of actions, that is part of what an SEO audit covers. Our own hours, directions and reviews are on AmpLocal’s Google Business Profile.
Frequently asked questions
Can I see AI search visibility in Google Analytics?
Not reliably. Some assistant referrals appear as traffic sources, but interactions that end without a click, or end in a phone call, leave no record at all.
How often should I run the audit?
Quarterly is enough. Answers vary between sessions, so checking weekly produces noise rather than insight. Keep the question wording identical each time.
Are AI visibility tracking tools worth paying for?
They can be, once the fundamentals are done and you need trend data at scale. Understand that they measure how often you appear across sampled prompts, not a ranking position.
What is the best free signal that this is working?
Branded search volume and calls from your Google Business Profile. Both tend to move when more people are being told your name, even when no website visit occurs.
What if an assistant states something incorrect about my business?
Find the source of the error, which is usually an outdated listing or an old page still online, and correct it there. Correcting the assistant directly is not an option available to you.
