The short version
We ran 145 real buyer questions about Quince, the affordable-luxury clothing brand, through ChatGPT, Google Gemini, and Anthropic Claude. Then we read all 375 answers and traced every source behind them. A single visibility number would have hidden almost everything interesting.
- AI doesn't just count you, it reviews you. The models praised Quince's "elevated essentials at accessible prices" and, in the same run, repeated a third-party doubt about its sustainability claims.
- The three models don't agree. ChatGPT almost never brings Quince up, but when it does it ranks the brand near the top. Gemini brings it up constantly, then buries it deep in the list. They read different corners of the web, so they reach different verdicts.
- Even a big, content-rich brand has blind spots. AI never connected Quince to things its own customers keep asking about, like postpartum pumping and flattering fit.
What do those questions look like in real data?
Buyers don't type "women's clothing" into AI. They describe a situation and a doubt, and in real data those questions range from open "best in category" searches to head-to-head duels to skeptical brand checks. The variety is the point.
Pulled straight from the Quince report, the shapes look like this:
- Awareness: "Best postpartum capsule wardrobe clothing brands for a pear-shaped figure that hide belly on a budget"
- Consideration: "What brands sell directly to customers without going through retailers, and are the prices actually better that way?"
- Decision: "Are Quince's organic clothing claims actually certified, or is it just marketing hype?"
- Head-to-head: "Eileen Fisher vs Quince for sustainable materials"
One whole cluster was about new parents: nursing-friendly clothes, postpartum jeans for wide hips and a smaller waist, underwear that won't irritate a C-section incision. Nobody guessing keywords for a clothing brand writes "C-section incision." Real buyers do, and the report was full of them.
What does AI actually say about Quince, good and bad?
Here is what the three models said about Quince across 375 answers, run with live web search on. Most of it was positive: 218 answers leaned that way, 82 stayed neutral, and 22 turned negative. The tally matters less than what sits inside each bucket.
The praise is specific. Claude called Quince's lineup "elevated essentials in luxurious fabrics at accessible prices" and said its "activewear exceeded expectations for its price point." That is the affordable-luxury story landing exactly as intended.
The doubts are just as specific, and they come from sources AI trusts. Asked whether Quince's organic claims are real, Gemini answered that there are "significant gaps between their claims and verifiable actions." On sustainability, it cited Eco-Stylist, which found the claims "not fully substantiated." A shopper who asks that question reads this answer, shaped by a reviewer Quince may not even know is in the conversation.
Then the blind spot. Across 72 open "best in category" questions, the kind a brand-new customer asks, AI named Quince in just 2. Ask "which brands make really nice quality clothes without charging luxury prices," almost the exact pitch Quince is built on, and it often doesn't come up at all. Quince gets a good review when it shows up. It rarely shows up at the moment a new customer is deciding who to even consider.
Do ChatGPT, Gemini, and Claude even agree about Quince?
The three models do not agree about Quince, because each one learns about it from a different corner of the web. This is the finding a single number erases completely.
Look at how differently they behave. ChatGPT named Quince in only 2.9% of its answers, but when it did, it placed the brand near the top, around position 2, and spoke the most warmly of the three. Gemini was the opposite: it named Quince the most, 11.4% of the time, yet buried it around position 24 and was the most lukewarm. Claude sat in the middle on all counts.
Why such different verdicts? Because they are reading different things. Gemini leaned on Reddit, with 51 citations, plus YouTube and Business Insider. Claude leaned on Quince's own site, with 138, alongside comparison and editorial sites like Apartstyle and The Good Trade, and the same critic, Eco-Stylist, that fed its sustainability doubts. ChatGPT's sources were thin and scattered: a few from quince.com, a couple from J.Crew, even a ThredUp resale page. Three models, three evidence bases, three opinions. Your AI visibility is not one conversation. It is three.
What are Quince's customers asking that AI doesn't connect to Quince yet?
Quince's customers are already asking about things AI never connects to the brand, like flattering fit and postpartum pumping. People are clearly asking for it, and Quince's name just doesn't come up.
For all its size, there were attributes AI never tied to Quince once: innovative design and flattering fit, hands-free pumping design, heritage and craftsmanship, all at zero mentions. Yet the questions buyers asked were full of exactly that, postpartum bodies, pumping, fit for a pear shape. AI hears the demand constantly and doesn't think of Quince. A few more attributes barely registered, mentioned only in passing: ethical manufacturing, price transparency, size inclusivity. Each gap is a chance to learn what your customers want and earn the coverage that teaches AI to connect you to it.
Why we built Adacity as a question-discovery engine
We built Adacity question-first because the questions are the hard part, and they have to fit who you are, not a keyword. Everything above came from starting in the right place.
We build the questions from a corpus of real consumer conversations, then add your own site, what social platforms say about you, and the sources your category trusts. We phrase them the way buyers actually ask. When a brand makes a claim, we ask the question a skeptical shopper would ("is this just greenwashing?") instead of testing the marketing phrase ("sustainable materials"). Then we run them across ChatGPT, Google Gemini, and Anthropic Claude with live web search on, read every answer, and trace every source. A single Deep Report runs 150 to 400 or more questions. You get the questions, the answers, and where they came from, not a spreadsheet to maintain.
How is that different from other AI visibility tools?
Most other tools charge a monthly subscription to track a short, capped list of prompts you pick yourself. You decide what to track, you guess at the wording, and you keep the list current.
Start with price. Profound runs $99 to $399 a month, AthenaHQ starts around $295, Scrunch at $250, Otterly from $29 to $489, Videntic roughly $230 to $1,050, and Evertune is enterprise-only, reportedly near $3,000 a month. Adacity is $10 to $39 per report, with no subscription.
Then depth. Entry plans cap what you can track. Otterly's cheapest tier follows 15 prompts, Profound's starts at 50, Scrunch's at 125. We ran 145 in a single report.
Last, where the prompts come from. Scrunch's own guide tells you to build them from your SEO and paid-search keywords. Otterly's research tool starts from your keywords and a URL, then asks you to choose. To be fair, some tools do suggest a starting set, Profound and Evertune among them. The gap isn't that nobody tries. It's that a list seeded from your industry or your keywords still isn't built from how a new customer actually phrases the question. That is the difference between measuring your marketing language and measuring your customer's, a distinction we dug into in what the Comscore AI report measures.
See what AI says about your business
Your customers aren't typing keywords into AI. They're asking real questions, and AI is answering them right now, with or without you. Want to see the questions AI gets about your business, the answers it gives, and where those answers come from? Run the $10 check. You get your questions, your competitors, and what to improve, in your inbox within the hour.