How we measure whether AI search is actually working for you
AI search is new enough that the metrics around it are full of noise, and plenty of vendors will happily sell you a number that sounds impressive and means nothing. Reston is a market where your customers genuinely use ChatGPT, Google's AI Overviews, and Perplexity to shortlist businesses before they ever click a website — so measuring this right matters. Here is what we actually track, and what we ignore.
The metric that matters most is simple: when someone asks an AI assistant a real buying question in your category — "who does bathroom remodels in Reston," "best family dentist near Reston Town Center" — does your business get named, and is the description accurate? We test this the way your customers do, running the prompts a real Reston buyer would type across the major assistants, and we log whether you appear, how you are described, and who gets named alongside you. That is the scoreboard. Everything else is supporting evidence.
The second real signal is referral traffic from AI sources. Assistants increasingly link out, and those clicks show up in your analytics as referrals from ChatGPT, Perplexity, and Google's AI surfaces. This traffic is small today but tends to be unusually high-intent — someone who clicks through after an AI already recommended you is far down the decision path. We watch whether it is growing and whether it converts, not just whether it exists.
The third thing we measure is accuracy and citation. Being mentioned is worthless if the AI has your hours, service area, or specialty wrong, so we track whether assistants are pulling correct facts and whether they cite sources we control — your own pages — versus a stale third-party listing. When an AI is citing a directory instead of you, that is a fixable problem we can see and act on.
- Tracked: whether AI assistants name you for real buying prompts in your category
- Tracked: AI-referred traffic and whether it converts into calls and forms
- Tracked: whether assistants describe you accurately and cite your own pages
- Ignored: a single "AI visibility score" with no methodology behind it
- Ignored: raw mention counts for searches no customer would ever type
What we deliberately ignore is any vanity dashboard that spits out one blended "AI score" with no way to see how it was calculated, and any report that counts mentions for prompts nobody in Reston is actually typing. A number you cannot trace back to a real customer question cannot tell you whether the work is paying off. We would rather show you the exact prompts, the exact answers, and how they changed month over month — the same plain-English standard we hold everything to. If you want to see where you currently stand across the major assistants, that is the first thing a conversation produces.