What It Costs to Be Invisible in AI Search

Being invisible in AI search is not a marketing problem. It is a revenue line you cannot see. A few founders have asked me lately for the number. What does it cost a tasting room when someone asks ChatGPT or Gemini where to go wine tasting this weekend and your name never comes up? They want a figure they can put in a slide, and I understand the instinct. I am going to disappoint you on purpose, because the honest answer is that no one can hand you that number without inventing it.

That does not mean the cost is zero. It means the cost is hiding inside your own numbers, not in someone else's headline statistic. So instead of selling you a fear, I want to walk you through a method you can run on your business this week, using figures you already have. By the end you will have a defensible estimate of your own exposure, built entirely on your own data, with not one borrowed number in it.

Can anyone tell you exactly what AI invisibility costs?

No, and you should be wary of anyone who claims they can. Any precise dollar figure for lost AI revenue, dropped onto your business from the outside, is essentially a guess dressed up as a fact. I have seen the eye-catching percentages floating around, and I will not repeat them here, because I cannot verify a single one against your tasting room. A made-up number that happens to scare you is still a made-up number, and using it would make me exactly the kind of advisor I tell founders to avoid.

The better approach is the one a good CFO uses on any uncertain line. Stop reaching for an industry average and build the estimate from the inputs you actually control. You do not need a precise market statistic to size your own exposure. You need a few numbers off your own books and one honest assumption you are willing to defend. That is a far stronger foundation than a downloaded chart, and it is the only version of this math I am willing to put my name on.

How do you size it with your own numbers?

Start with this honest question. Of your first-time customers last year, how many came from a recommendation? I am not referring to an ad they clicked or a coupon they redeemed, but an actual recommendation. A friend who said you have to go here, a list someone found, a name that came up when they asked where to go. For many tasting rooms and local brands, that share is significant and continues to grow. You may not have it to the decimal, but you know roughly whether it is most of your new faces or a sliver of them.

Next, consider where those recommendations are starting to come from. A friend's suggestion still matters and always will. But more trip planning is quietly shifting to AI assistants, much like the transition from guidebooks to Google. You do not need a precise percentage of that shift to feel the exposure. You only need to accept that a growing share of the people who would once have asked a person or a search engine are now asking an assistant, and that the assistant either names you or it does not.

Now put a value on a single recommendation. Take the average value of a first-time customer and multiply it by what they are worth over their first year, factoring in those who join a club or return for another visit. Not the price of one tasting flight. The first-year value of a relationship. That calculation determines the value of one recommendation that actually lands a customer at your door.

Then connect the two. Each list you are not included on, every time an assistant answers that weekend-plans question without your name in it, represents a fraction of that first-year value walking past your driveway. One miss is small. The same miss, repeating quietly throughout the year across every person who asks, is not. You now have a range, built from your recommendation share, your first-year value, and your own honest sense of how often you are in the answer versus left out of it. No invented figures required.

Why doesn't this cost show up anywhere?

It does not show up because it is revenue that never arrived, and your accounting system can only count things that happened. A discount shows up. A refund shows up. A marketing invoice shows up. But a customer who asked an assistant for a recommendation, never heard your name, and went somewhere else generates no transaction, no line item, and no trace in your books. The P&L is a record of what occurred, and a non-event occurs nowhere.

That is what makes this gap so easy to ignore. A cost you can see is a cost you can argue about, cut, or defend. This one produces no document. There is no monthly statement titled customers who never found you. So it sits outside the entire apparatus your business uses to decide what matters, which is precisely why it deserves a deliberate look rather than a hope that it is small.

Why does a finance person care about a marketing-looking gap?

Because it is not a marketing gap. It is a finance gap wearing a marketing costume, and this is the part that concerns me. It is not a visible cost you can cut. It is revenue that simply does not materialize, goes unaccounted for, and therefore is not advocated for by anyone. A visible cost has an owner. Somebody is responsible for the ad budget, the discounting, the staffing line, and somebody will fight to defend or trim it. Revenue that never arrives has no owner.

This is the same instinct behind why your CAC number leaves out the most expensive part of acquisition. The numbers that quietly shape your business are usually the ones missing from the report, not the ones at the top of it. A finance person earns their seat by naming the line nobody else is looking at and asking what it is worth before deciding to ignore it. Invisibility in AI search is exactly that kind of line. Large enough to matter, quiet enough to escape every report you currently read.

What does this mean for you?

It means you do not have to speculate about your exposure, and you also do not have to panic about it. You can run the math above on your own numbers in twenty minutes and come out with a defensible range, not a borrowed scare. Maybe that range is small for you, and you file it and move on with a clear conscience. Maybe it is large enough that it changes where you spend your next marketing dollar. Either way you will know, and knowing is the entire point.

What you should not do is assume the answer is zero just because nothing in your books flags it. Absence of a line item is not absence of a cost. The same logic applies whether you run a tasting room, a wellness studio, or a local brand, which is why this is core to how we think about helping wineries and other owner-led businesses see the numbers their software will never surface on its own. The work is not to invent a statistic. The work is to look honestly at the one cost your reports were never built to show.

Your Monday-morning next step

Monday morning, before you run any of that math, find out whether you have a problem at all. You can do the exercise on a napkin, but the missing input is the simplest one. When someone asks an AI assistant about wineries like yours, does your name come up? You do not have to speculate about your own exposure. We built a free snapshot that reveals what the major AI engines currently say about your winery, with no sales pitch involved. It does not promise to fix anything or move you up any list. It simply shows you, in plain terms, what the major engines currently say when a prospect asks, so you are not guessing about the one number that anchors all the others.

Start there, then run your own math. And once you have it, sit with the question that started this whole exercise: what would one more recommendation a week be worth to your business?