Shoppers ask ChatGPT and Perplexity what to buy, and get back a list of brands. We find out whether yours is on it, why not, and which half of the problem can actually be fixed.
Free check takes two business days. No card, no call required.
Being absent from AI answers is usually one of these. Some are mechanical and we fix them in weeks. Some are slower than that, and we label them rather than sell them to you.
A line in robots.txt tells ChatGPT, Perplexity or Gemini not to read the site. Usually added years ago during a scraping panic, or switched on by a CDN default nobody chose. Nothing else can be fixed until this is.
The one defect that makes every other fix pointless
No AI crawler runs JavaScript. If your product details are drawn in by the browser after the page loads, assistants fetch the page and find an empty shell. Passing a Google check proves nothing here: Google renders, they do not.
Invisible to every standard SEO tool
Product pages with no machine-readable price, stock status or identity. This matters for the assistants that read your site live and for shopping surfaces. It is not what makes a model recommend you from memory, and we do not claim otherwise.
The most common mechanical finding by a wide margin
Shopping results draw on merchant feed infrastructure. A store that never submitted a feed is structurally absent from it, no matter how good the site is. Where a feed exists, the eligibility flags are often silently switched off.
Almost never set up outside large retailers
No sitemap, no feed, no crawlable product URLs. Even when the brand is known, individual products cannot be surfaced or recommended.
Common on custom and older builds
Specific buying questions where the brand never appears and the same four competitors always do. We show you the exact questions and the exact names. This is an authority gap, not a markup gap: worth knowing, slower to change.
The finding that changes how owners think about this
Named eleventh out of fourteen brands. Technically in the answer, practically invisible: nobody reads to the bottom of a recommendation list.
Invisible to any check that only asks "did we appear?"
Everything technical is clean and the brand still does not get named. We have measured stores with perfect product markup scoring zero. That is an authority problem, and pretending markup will fix it would waste your money.
The finding most agencies quietly misdiagnose
The free check tells you, with the actual questions and the actual competitor names. Two business days, no card, no call.
We ask the AI systems your customers use the questions your customers ask. No mention of your brand: a result only counts if the system names you on its own. You see exactly what came back, free.
Every question, every competitor named, every technical defect, ranked by what it costs you. Split into what we can fix and what no agency can fix quickly. Written so you can hand it to a developer or read it yourself.
Crawler access, rendering, product schema, merchant feeds, identifiers, catalog discoverability. The work that moves you from unreadable to readable, verified before and after, so you can see it happened.
These systems change their rules without announcing it, and feeds break silently. Weekly re-checks, and you hear from us when something moves.
The exact structure every check produces: the same categories, the same question types, the same technical checks, whatever you sell. This is the document, not a mock-up of one.
AI shopping visibility report
Every check follows this structure
Typical turnaround: 2 business days · anthropic, perplexity, chatgpt, gemini
Visibility score
? / 10
Out of 10, based on how often your brand is named across unbranded buying questions.
The score splits in two
Does the model know you
Being written about widely. Slow, months to years.
Can it read your site
Technical fixes. Weeks, and verifiable.
What we check
Who shows up instead of you
Every check ranks the brands named in your answers, illustrated here with placeholder names.
The kind of question every check asks
anthropic What are the best [category] to buy right now?
who gets named: Whichever brands the model already trusts, named or not
perplexity Where should I buy [category] online?
who gets named: Retailers and brands with a live, readable product feed
gemini Which brands make good [category]?
who gets named: Brands the model has read about widely
14 questions per check, written for your category
14
unbranded buying questions per check
4
AI engines tested, per question
3
technical categories audited
Shown as a template, not one client's numbers. Every check runs against your own brand, your own competitors, and your own catalog.
Most of this industry is a year old and a lot of what gets sold in it does not survive contact with the evidence. Here is what we measure, and what we refuse to promise.
Ask an assistant the same question twice and you can get two different brand lists: fewer than one in a hundred repeated prompts returns an identical list. Anyone quoting you a score from a single run is quoting noise. We repeat and report the spread.
Some assistants read your site when they answer. Others answer from training data and never open it. Technical work moves the first group and does nothing for the second. Every report says which is which, per engine, instead of blending them into one flattering number.
Nobody controls what a model says, and any agency guaranteeing placement is guessing. We commit to specific technical outcomes: readable pages, a live feed, unblocked crawlers. We show you the before and after.
The free check exists to find that out. Some stores are already visible and already readable. When that is the answer, we say so and there is no pitch attached.
Start with the free check. Nothing else is worth buying until you have seen whether there is a gap worth closing.
Visibility check
Freeno card, no call required
We run your brand through the AI systems shoppers use and send you what came back.
Full audit
$650USD, invoiced after the call
Book a call first. We scope it live, you get a fixed price, and you only pay if you go ahead.
The complete picture across every AI surface, split into what we can fix and what we cannot.
Fix package
$1,800+USD, scoped per catalog
Half to start, half once the fixes are verified live.
We do the work: crawler access, structured data, feeds, identifiers.
Then, to stay visible
Monitoring
$325USD, per month
Weekly re-checks, because these systems change and feeds break quietly.
All prices in US dollars. The audit and fix package are scoped on a call and invoiced afterwards, no card up front, and no charge if you decide against it. Monitoring runs through Stripe, so card details never touch this site.
These are the ones we actually get asked, including the sceptical ones. Answered straight.
Ask ChatGPT "best sleeping bags under $150" and you get a list of brands, not a page of links. Roughly 50 million shopping questions a day get answered that way. Nobody chose which brands appear on those lists: the systems assembled them from what they could read and what they had been trained on. That is the whole thing.
The first check is free, takes no card and no call, and it is your own store being checked. You get the questions we asked, the answers that came back, and the names of whoever appeared instead of you. If that is not useful you have lost nothing and we have not asked you for anything.
We cannot promise a dollar figure and we will not pretend otherwise. The channel is too new for anyone to have honest numbers, and any agency quoting you a revenue lift is guessing. What we can prove in five minutes is whether shoppers asking about your category are being shown your competitors and not you.
No, and neither can anyone else. Nobody controls what a model says. We commit to specific technical outcomes instead: crawlers unblocked, product data readable, a live merchant feed. We verify each one before and after so you can see it happened.
Possibly, but this is not standard SEO work and most of it is invisible to standard SEO tools. Take the free check and hand them the findings. If they have already covered it, the report will say so and that is a good outcome.
You do nothing. We check from the outside: no access to your site, no logins, no install. You get a report written so that you can read it yourself or forward it to whoever does the technical work.
Small stores are the ones being left out right now. Large retailers already have people on this. In practice the brands with the biggest gaps are the ones nobody has written about much yet, which is most independent brands.
Then we tell you that and there is no pitch. It happens: some of the brands we have checked score above nine and genuinely do not need us. The free check exists to find out which group you are in.
Depends which half of the problem you have. Technical readability is weeks and we can verify it. Whether a model knows your brand at all is months to years and no agency can shortcut it. Your report separates the two so you know what you are actually buying.
Three fields, no card, no call. We run your brand through the AI systems shoppers use and send you what came back.
Book a call instead