Freevine Labs LLC is based in Dallas, Texas, and was founded by Calvin Li and Lisa Li. We help small and midsize businesses win customers in the AI era: we build websites and get them found on Google Search, on Google Maps, and in AI answers like ChatGPT’s, then keep improving them based on our public AI Visibility Index and Signals data, so traffic and customers compound over time.
Lisa and I started Freevine Labs together. In seven years of running businesses in North America, one question kept coming back: once the product and the service are good, where does the next customer come from?
What stayed with me from those years is the trade-off every small business owner faces daily: take care of the customer in front of you, or stop and chase the latest change in how search works. Now customers ask ChatGPT who’s good near them. The front door has moved, and there’s a new gap between doing good work and being seen.
So we decided to measure that gap first. Freevine asks a fixed set of questions, in English and Chinese, in the same place, on the same model, in the same window, again and again, and records who ChatGPT actually names. Whatever clears our publication threshold goes into the AI Visibility Index (currently in Chinese).
When a business is ready to improve its website and its listings, we turn those observations into work that gets done. No shortcuts, and no promises we can’t keep. We just don’t want a good local business to lose its next customer because AI hasn’t learned to read its website yet.
Calvin LiCo-founder, Freevine Labs · Dallas
How we measure (Chinese)City by city and industry by industry, the index shows how often local businesses come up when ChatGPT is asked a fixed set of questions. Independent samples are averaged first, then question wordings and intents are weighted equally. Citations, Google evidence, and website evidence are shown side by side, never blended into one score.
For small and midsize businesses we build websites Google and AI can read and run AI search optimization (SEO and GEO), with the public index and our own industry database setting the priorities. Scope, prices, and acceptance criteria are published on every service page.
Every release repeats the same questions, in the same place, on the same model, in the same time window. If the valid sample is too small to publish, we don’t publish a percentage.
Claiming a listing, requesting a correction, asking a question, buying a service, or sponsoring a placement never changes AI visibility or rank.
Search results are checked in the client’s own accounts, so every number in a report traces back to where it came from.
How we define the metrics, draw the sample, version each release, and handle corrections is all in theAI visibility methodology (Chinese)
Want to check an index record, or talk about your website and search? Reach us directly.
