How we measure whether AI search is actually working for you
AI search is new enough that the industry is still inventing metrics for it, and most of what gets reported is noise dressed up as insight. Here's how we actually tell whether the work is moving the needle in Staunton — and which numbers we deliberately ignore because they look impressive and mean nothing.
The metric that matters most is inclusion: when a real buyer asks ChatGPT, Gemini, or Google's AI overview a question your business should answer — "best HVAC company in Staunton VA," "who does roof repair in Augusta County," "good restaurant near Blackfriars Playhouse" — do you show up in the answer at all? We build a list of the questions your actual customers ask, test them on a schedule, and track whether your business appears, how it's described, and whether the description is accurate. Going from absent to mentioned is the single clearest signal the work is landing.
The second thing we watch is how you're framed. Being mentioned isn't enough if the AI describes your category wrong, lists services you don't offer, or attaches stale information. We track the substance of what's said about you, because a customer acts on the description, not just the name. Fixing a wrong or thin description often matters more than adding one more mention.
The third, and the one owners feel in their bank account, is downstream: are you getting inquiries where the person already knows more than they should from a cold Google search? When someone calls and references specifics — your service area, a price range, a detail about how you work — that's often an AI tool having done the research and handed them to you pre-qualified. We ask new customers how they found you and watch for that pattern over time.
Now the vanity metrics we ignore. We don't chase a made-up "AI visibility score" that no platform actually publishes — those are marketing inventions, not measurements. We don't count how many times your name appears across a thousand irrelevant prompts nobody in Staunton would ever type. And we don't celebrate being mentioned in an answer to a question that never leads to a sale. Volume for its own sake is easy to manufacture and worthless to your business.
- The measurement is honest about scale. AI search is a real and growing channel here — a genuinely small, discoverable market where visitors constantly ask AI tools for recommendations makes it worth working now — but it is not yet your biggest source of customers, and we won't pretend a handful of mentions is a flood. We report the real number.
The through-line is simple: we measure what a customer would actually act on, and we throw out everything designed to look good in a report. If you want to see how the AI tools describe your category in Staunton today, before spending anything, that's where we start — with a plain read of where you stand.