The Fashion Brands ChatGPT Recommends Most (5,072 Carousels Analyzed)

Among the fashion brands ChatGPT recommends, Lulus leads: it appears in 1,515 of 5,072 product carousels, ahead of Petal & Pup (1,208) and Fashion Nova (844). The winners are mostly DTC brands that own their product feeds and publish clean prices and complete attributes. Fame doesn't decide it: Zara appears in just 100 carousels.

Last updated: October 17, 2026. Data as of: February 28 and March 1, 2026 (AEOsome Research, 2026). We ran 8,520 real shopping queries through ChatGPT (women's dresses, US market, served by gpt-5-mini) and recorded all 5,072 product carousels that appeared. For each brand we logged appearances, price, and rating. We ran the queries ourselves and computed every count directly from the raw JSONL export, not from a third-party tool. The collection dates, the model and the query design are documented in the study methodology chapter.

TL;DR: Brands that own their product pages and keep prices and attributes complete dominate ChatGPT's dress carousels. Famous brands do not get in on fame alone. Average rating alone does not explain who wins.

Three terms matter here. Share of voice (also called carousel presence) is the number of unique carousels in which a brand appears at least once, divided by all 5,072 carousels. A product carousel is the row of up to 8 shoppable products ChatGPT shows inside an answer. A citation is a source link ChatGPT embeds in the response text.

ChatGPT strongly favors DTC brands with direct-to-consumer product feeds. Department stores appear as product hosts, not as recommended brands.

Our wider study agrees: in women's dresses, 42.3% of card slots sit on brand-owned stores, against 2.2% to 6.6% in the other categories we tested (AEOsome Research, 2026, Study 2, n = 22,335 de-duplicated answers; see the category chapter). ChatGPT's own shopping tool is built on a shopping-tuned GPT-5 mini variant and pulls from public retail pages, according to Retail Dive's report on OpenAI's launch.

Key findings from 5,072 ChatGPT shopping carousels:

• Lulus appears in more carousels than any other brand: 1,515 out of 5,072
• Fashion Nova ranks #3, appearing in 844 carousels at an average price of $32
• Zara has 237 confirmed appearances across 100 carousels. It's visible, but dramatically underrepresented for its market size
• Alibaba is the single most-cited domain in ChatGPT shopping responses, ahead of Nordstrom, Lulus, and every US retailer
• Quince ranks #6 overall with an average customer rating of 4.73 stars, the highest of any top-10 brand and tied with Express (#10)

These are counts from the raw JSONL dataset, not estimates.

In this article:

• Methodology: How We Counted Brands
• The Brands ChatGPT Recommends: The DTC Leaderboard
• Retail Platforms: Where ChatGPT Sources Products
• The Surprises: Brands Nobody Expected to See Here
• The Absences: Major Brands ChatGPT Is Not Recommending
• The Citation Layer: Which Domains ChatGPT Sources From
• What Separates the Brands That Win
• How to Improve Your Brand's Visibility in ChatGPT Shopping Results
• Limitations
• Frequently Asked Questions
• Summary

Methodology: How We Counted Brands

Most published analyses skip this distinction.

Brand presence counts how often a product linking to a brand's own domain appeared in a carousel. A Lulus dress whose product URL points to lulus.com is a confirmed Lulus appearance. Lulus made the product and owns the page.

Retailer platform presence is a separate metric. A Karen Kane dress whose URL points to macys.com counts toward Macy's as a platform. Macy's hosts that product; it isn't the brand. Calling it a "Macy's brand appearance" would be wrong.

The tables keep the two apart.

The Brands ChatGPT Recommends: The DTC Leaderboard

These brands' products link to their own domain, so each controls its product page, data, and feed.

Note: This dataset was collected on February 28 and March 1, 2026, and covers Women's Fashion > Dresses queries only, US market, on ChatGPT (gpt-5-mini). Rankings reflect share of voice within this category and query set. "Appearances" = total product slots across all triggered carousels. "Carousels Present In" = unique queries in which the brand appeared at least once.

Retail Platforms: Where ChatGPT Sources Products

ChatGPT also surfaces products from department stores and marketplaces. These are product hosts; the brands they sell may be other companies.

What this means for brands: Listing on Macy's, Nordstrom, or Etsy lets ChatGPT show your products even when it isn't citing your domain. That's a multiplier, not a substitute for owning your product feed.

The Surprises: Brands Nobody Expected to See Here

The DTC brands nobody saw coming

Lulus (#1), Petal & Pup (#2), Fashion Nova (#3) and Quince (#6) all link to their own domains, and they rank above H&M (#9) and Gap (#14).

Lulus (#1) began as a vintage boutique in 1996, started selling online in 2005 and has been purely online since 2008, according to its SEC Form 10-K. It appears in nearly 30% of all triggered carousels, more than any department store, fast-fashion giant, or brand with physical stores.

Petal & Pup (#2) is an online-first DTC brand. It ranks above every specialty retailer except Lulus, at a mid-market $93 average.

Quince (#6) sells through a direct-to-manufacturer model. It averages $76 per dress and carries a 4.73-star rating, tied with Express for the highest of any top-10 brand. This dataset does not show why ChatGPT selects Quince, so we make no claim about its product data.

Fashion Nova (#3) ranks above Abercrombie, Gap, LOFT, and Ann Taylor, covering a $32 budget tier mid-market brands can't match.

The ratings paradox

H&M ranks #9 with the lowest average rating of any top-10 brand (4.21★). ASOS ranks #19 with 4.20★. Both are global brands.

Average rating does not decide which brands get in. Lulus is #1 with a 4.29★ average, while Express (4.73★) is #10, Ever-Pretty (4.76★) is #13 and Rihoas (4.86★) is #24. Our wider study found the same thing: across 110 brands with at least 50 slots, average rating is negatively correlated with the number of carousels a brand appears in (Spearman −0.24; AEOsome Research, 2026, hub study). Rating matters after a brand is in the carousel: within an eight-slot carousel, one standard deviation of rating doubles the odds of rank one (odds ratio 2.1). We unpack this in reviews as a ranking signal.

Some very highly rated brands still sit low or mid-table: Oh Polly (4.84★, rank #21), Rihoas (4.86★, rank #24), and MESHKI (4.87★) all appear in carousels, and none is a household name in US fashion retail.

The Walmart efficiency story

Walmart appears as a retail platform for 1,672 product slots across 827 carousels. At an average price of $23, it covers every budget-framing query in the dataset. Walmart and Old Navy ($23 average) form the low-price tier, where price is the main sorting signal.

The Absences: Major Brands ChatGPT Is Not Recommending

This is the finding that should worry major fashion retailers most.

Zara appears 237 times across 100 carousels. It's present, but in fewer than 2% of all triggered carousels, against 30% for Lulus. Zara's group (Zara, Zara Home and Lefties) reported €28.05 billion in sales for fiscal 2025, so that is severe underrepresentation.

SHEIN appears 31 times, in 23 carousels. Amazon Fashion appears about once.

These are global fashion's dominant players. With 5,072 carousels in the sample, the gap looks structural rather than a sampling anomaly, though we have not re-run it since February 2026.

One hypothesis, which this dataset cannot confirm: product feed gaps and schema coverage. ChatGPT's shopping tool pulls from public retail pages (per Retail Dive, above), so brands with inconsistent or underattributed SKU-level data may surface less often, however famous offline.

The Citation Layer: Which Domains ChatGPT Sources From

Citations tell a different story than carousels.

Alibaba is the most-cited source in ChatGPT fashion shopping responses, by a wide margin.

wedding.alibaba.com alone generates 4,543 citations across 2,102 unique queries. Add alibaba.com (4,047 citations) and the Alibaba group reaches 8,590 total citations. That's more than three times nordstrom.com (2,213).

Alibaba is a wholesale marketplace, not a brand. A plausible reason it is cited so often (a hypothesis, not tested here) is that it hosts structured listings at massive scale, with machine-readable prices and attributes.

nordstromrack.com is cited more than nordstrom.com (2,557 vs 2,213). The off-price Rack outdraws the flagship, likely reflecting price-sensitive queries.

reddit.com is #8 with 1,933 citations across 1,352 unique queries. A UGC platform is cited in shopping contexts more than target.com (1,233) or dillards.com (1,183). This may mean ChatGPT treats Reddit discussions as a source of shopper opinion, though this dataset does not test why.

lulus.com is #10 for citations (1,839), consistent with its #1 carousel position. Lulus commands both surfaces at once.

Some domains you might expect near the top sit much lower. To see how the other engines compare with ChatGPT on the same questions, read our four-engine comparison.

• poshmark.com: 442 citations (not in top 15)
• macys.com: 439 citations (not in top 15)
• lyst.com: 134 citations
• depop.com: 29 citations

What Separates the Brands That Win

Five structural patterns explain why the top 25 appear and the absent brands don't.

Pattern: Ratings help ranking inside the carousel, not entry

Why it works: Across 110 brands, average rating did not predict how many carousels a brand appeared in (Spearman −0.24, AEOsome Research, 2026), but inside a carousel higher rating raised the odds of rank one (odds ratio 2.1 per standard deviation).

Example: Quince (4.73★, #6) outranks H&M (4.21★, #9), yet Lulus (4.29★) leads the whole table.

Pattern: Price signal clarity

Why it works: Budget Framing queries ("best dress under $100") trigger carousels 88.8% of the time, one of the highest rates of any query type. Brands with clean, accurate, consistently structured prices are a plausible match for these queries.

Example: Fashion Nova ($32 avg) and Old Navy ($23 avg) both rank in the top 4.

Pattern: Broad price range coverage

Why it works: Spanning several price tiers can match more queries than one band.

Example: Anthropologie ($157 avg) and Adrianna Papell ($163 avg) still make the top 16, alongside budget brands.

Pattern: Attribute completeness at SKU level

Why it works: Attribute Constrained queries ("floral, knee-length, cotton dress") trigger carousels 81.8% of the time. Products with complete attribute data (material, occasion, fit, care instructions) are the natural match for these queries. Google's product data specification lists the attributes merchants are expected to supply.

Example: Petal & Pup ranks #2 (1,208 carousels). We did not audit its product pages, so attribute density is a hypothesis here, not a finding.

Pattern: Direct product feed ownership

Why it works: Brands that control their own product URLs can keep data consistent and current. Hosting third-party products on a department-store page may add data lag, though we did not measure it.

Example: Lulus, Fashion Nova, Old Navy and Quince all link to their own domains. We did not audit the quality of their feeds.

How to Improve Your Brand's Visibility in ChatGPT Shopping Results

1. Audit your average product rating. Our data shows rating does not get a brand into more carousels, but once a product is shown, a higher rating raises its odds of the top slot.
2. Check your price data for completeness. Every SKU needs a clean, current price. Missing or stale prices exclude products from budget-constrained queries.
3. Expand your attribute coverage. Add material, occasion, fit type, and care instructions to every product page. These fields match Attribute Constrained queries (81.8% carousel rate).
4. Maintain your own product feed. The top-ranked brands all link to their own domains.
5. Submit structured product data to aggregators. nordstromrack.com and windsorstore.com are top-10 citation domains. Listings there multiply your citation surface area.
6. Get listed on high-citation domains. nordstrom.com (2,213 citations), lulus.com (1,839 citations), and target.com (1,233 citations) act as citation amplifiers. A product listed there reaches ChatGPT through multiple URL paths.
7. Add JSON-LD product schema to every product page (see Schema.org Product). Structured data is a plausible lever, though this study did not test it.
8. Don't rely on brand recognition. Zara's limited presence (100 carousels out of 5,072) shows awareness alone does not translate into AI visibility. Structured data is the lever we recommend testing.
9. Monitor your AI share of voice by category. The leaderboard shifts as feeds update. Track your appearances against your category's top 5 brands monthly.

Limitations

• One category: Women's Fashion > Dresses.
• One model: ChatGPT served by gpt-5-mini, US market.
• One snapshot: February 28 and March 1, 2026, not re-run since.
• Brand presence is attributed by product URL domain, so a brand sold only through other retailers is counted under those retailers.
• The patterns above are associations in this dataset, not tested causes.

Written by Vijaya Kumar Channalli, AEOsome. Corrections and questions about the counts are welcome via aeosome.com.

Frequently Asked Questions

Which brands appear most in ChatGPT fashion recommendations?

Across 8,520 ChatGPT shopping queries in Women's Fashion > Dresses, Lulus appeared in the most unique carousels (1,515). Next came Petal & Pup (1,208), Fashion Nova (844), Old Navy (826), and Abercrombie & Fitch (751). All sell mainly direct-to-consumer, with structured, machine-readable product data.

Why does Zara appear so infrequently in ChatGPT shopping results?

Zara appears in 100 unique carousels (237 product slots), fewer than 2% of the 5,072 triggered. The likely cause is product data coverage. ChatGPT's shopping surface relies on structured product feeds, schema markup, and retailer integrations. Brands with incomplete SKU-level data surface less often, however famous.

What is the difference between brand appearances and retailer platform presence?

Brand appearances count a brand's products in carousels, measured by its own product URL domain. Retailer platform presence counts products a department store or marketplace (Macy's, Etsy, Nordstrom) hosted, whatever the brand. A Karen Kane dress on macys.com counts toward Macy's platform presence, not Macy's brand presence.

Why does Alibaba rank #1 for citations on ChatGPT?

Alibaba's two domains, wedding.alibaba.com (4,543 citations) and alibaba.com (4,047 citations), generate 8,590 combined citations. That's three times any other source. Alibaba hosts structured listings at massive scale, with machine-readable prices, attributes, and supplier data. Few US fashion marketers see it as a competitor. In structured data volume, it dominates.

Does a higher product rating improve ChatGPT visibility?

Only partly. Across 110 brands, average rating is negatively correlated with carousel count (Spearman −0.24, AEOsome Research, 2026), so a high rating does not by itself get a brand shown more often: Lulus is #1 at 4.29★, while Rihoas (4.86★) is #24. Inside a carousel, though, one standard deviation of rating doubles the odds of rank one (odds ratio 2.1).

How does product price affect AI shopping recommendations?

Price matters in two ways. Price-constrained queries ("dresses under $100") trigger carousels 88.8% of the time, so products in common budget brackets match more queries. Clean pricing is also a completeness signal: missing or stale prices make products less likely to appear. Fashion Nova ($32 avg) and Old Navy ($23 avg) rank in the top 4 partly for this reason.

Summary

• Primary finding: AI shopping visibility is driven by product data quality and feed ownership, not brand recognition. Zara underperforms while Lulus and Fashion Nova dominate. Average rating does not explain presence
• Top brand by carousel reach: Lulus (1,515 carousels), followed by Petal & Pup (1,208)
• Biggest underperformer: Zara (100 carousels), SHEIN (23), Amazon (about 1)
• Most-cited domain: Alibaba (8,590 combined citations), a wholesale marketplace, not a US consumer retailer
• DTC brands to watch: Quince, Petal & Pup, Baltic Born, Ever-Pretty, Oh Polly. All in the top 25 of the leaderboard
• Quick win: Get listed on high-citation domains (nordstromrack.com, nordstrom.com, windsorstore.com, lulus.com, target.com). Each adds citation surface area beyond your own site

About the author. Vijaya Kumar Channalli is a serial SaaS founder and the founder of AEOsome (Tartu, Estonia), which helps eCommerce brands get their products recommended when shoppers ask AI what to buy. Read how we write, source and correct posts in our editorial policy.

Last updated October 17, 2026. This research is based on 8,520 shopping intent queries run against ChatGPT (gpt-5-mini), covering the Women's Fashion > Dresses category in the US market. Queries spanned 9 intent stages and 5 query styles. Brand and retailer citation data came from 5,072 triggered product carousels. Brand presence is measured via product URL domain for brand-owned sites; retailer platform presence is measured separately. Citation counts reflect the actual citations field in the raw JSONL dataset. The dataset is a snapshot of ChatGPT's commerce behavior as of the run date (February 2026).