How does DigiSnax measure whether AI cites you?
Updated October 11, 2026
Most companies selling AI visibility will tell you what they do to your website. Very few will tell you how they check whether it worked, because most of them do not check. We do, every month, for every client, and this page says how, within a limit: what we ask, where, how often and what we keep is below. Which questions, which pages, which tools respond first and how an answer is scored is the part we have spent the year tuning, and it stays with DigiSnax.
What we ask. We ask the kinds of questions a buyer or seller types before they have chosen an agent: the questions people ask about living in a community, and the questions they ask about its housing market. These are asked long before anyone types "best real estate agent near me," and they are the questions an agent's own website can become the source for. The work is being the answer before the hiring question is ever asked.
Where we ask. We ask AI tools that search the live web before they answer. That qualifier matters more than any count of tools. An assistant that answers from memory cannot cite anyone, and a measurement built on it would be measuring nothing. We ask from the client's own town, because the same question asked from somewhere else gets a different answer. And we ask each question more than once, because the same tool gives different answers to the same question on different days, so one answer tells you very little and a set of answers gives a rate. We report that rate with its range, so a client can see how firm a number is before leaning on it.
How often. Monthly, on a schedule, the same way every month.
What we keep. Every answer, every source each answer cited, and the date, stored as they came back. We record whether the client's website was among the sources an answer drew on, and whether the client's name appeared in the answer, because those are two different things and we should know both. Because the raw answers are kept, a reading can be checked later, and a client's history is never a number somebody wrote down once.
What the client sees. A monthly report, personalized and reviewed before it reaches them, in plain words: where their site was part of the answer, where it was not yet, and what moved since last month. The report does not carry the raw answers or the scoring; it carries what a busy agent needs to know.
What we will not claim. Nobody can promise a citation, and we do not. AI tools change what they read and how they answer, a market can go quiet for a season, and a new website is read by the tools weeks after it is built. What we can promise is that the number we show a client is a measurement with a known method behind it, taken the same way every month, and that we measure it whether it flatters us or not. We run the same measurement on our own website. A claim whose measurement a vendor cannot show you has nothing behind it that you can check.
Why publish this at all. Because an agent weighing DigiSnax's service, or any other, should be able to ask the vendor the same five questions and get a plain answer: what do you ask, where, how often, what do you keep, and what will you tell me when it does not work. Our answers are above, dated, and they will be updated if the method changes.
DigiSnax builds real estate websites engineered to be cited, and measures whether they are. If you want to see what AI says about your market today, tell us your market and we will show you.