How we tell whether AI search is actually working for you in Radford
AI search is new enough that plenty of people selling it lean on numbers that sound impressive and mean nothing. So before any of that, here's the honest part: the metric that matters is whether a real assistant — ChatGPT, Google's AI answers, Perplexity — names your business when a New River Valley customer asks it a buying question. Everything we measure is in service of that one outcome, and everything that isn't, we ignore.
What we actually track starts with direct answer testing. We run the real questions your customers ask — "best place for X near Radford," "who does Y in the New River Valley," "is Z open near campus" — across the major assistants on a regular schedule, and we record whether you're named, how you're described, and whether the details are right. That last part matters more than owners expect: an assistant that confidently recommends you with the wrong hours or an old phone number is a problem, not a win. Tracking the accuracy of how AI describes you is half the job.
The second thing we watch is citation presence — whether the sources these tools actually read (your own structured pages, your Google profile, the local and regional listings AI pulls from) say clear, consistent things about you. When an assistant recommends a Radford business, it's synthesizing those sources. If they're thin or contradictory, you don't get recommended, full stop. So we measure the quality and consistency of what the machines are reading about you, because that's the lever that moves the outcome.
Here's what we deliberately don't report, because in Radford these numbers flatter everyone and tell you nothing:
- Raw "AI impression" counts with no purchase intent behind them, follower or share counts on anything, generic traffic spikes that never turn into a call or a visit, and any single-day screenshot of you ranking that can't be repeated the next week.
The reason this discipline matters more in Radford than in a big city is that the field is genuinely open here, which makes it tempting to declare victory early. Being named once in one tool on one day is not the same as being the reliable default answer for your category across the New River Valley — and only the boring, repeated testing tells you which one you've got. If the honest measurements say it's working, you'll feel it in real calls and walk-ins, not in a vanity chart. That's the standard AI search work is held to, and it's the only one worth paying for.