I Audited UpCircle Beauty Against AI Shopping Agents. The First Brand That Actually Works.

Executive Summary

• Brand: UpCircle Beauty. UK-based sustainable skincare brand founded in 2016 by siblings Anna and William Brightman. Transforms by-products from the food and drinks industry into natural skincare. B Corp certified (score 107.2). Stocked at Ulta Beauty (768 doors), Boots, Holland & Barrett, Whole Foods, Credo Beauty. 600,000+ customers. Trustpilot 4.6/5 with 2,354 reviews.
• AI visibility score: 8/15 queries surfaced the brand. The highest score in this audit programme.
• The pattern: UpCircle appeared on every query that matched their sustainability identity. "Zero-waste moisturiser" — first on all three platforms. "Eco-friendly body scrub" — first on Gemini. "Coffee-based face scrub" — first on Gemini and Copilot. "Sustainable face serum" — first on Gemini and Copilot. But "rosehip face oil for dry skin" — a generic product-attribute query with no sustainability signal — zero on all three platforms.
• Key finding: Brand narrative translates to AI visibility, but only when the query contains narrative keywords. UpCircle owns the sustainability layer. They are invisible on the product-attribute layer. When someone asks for "a good face oil for dry skin," the upcycled ingredient story, the B Corp certification, and the 450 tonnes of repurposed waste do not help. The product data is too thin to compete on product merits alone.
• Root cause: The body_html descriptions are adequate but not rich. No INCI ingredient lists in the crawlable data. No usage instructions. No product comparison context. The Shopify product tags are among the best in the audit set (Sensitive Skin, Acne-Prone, Collagen Banking, Glass Skin, Double-Cleansing) but they exist only in the JSON API, not in page-level structured data.
• Fix complexity: Low. UpCircle is already winning the sustainability queries. The fix is extending that visibility into product-attribute queries by adding INCI lists, usage instructions, and skin concern context to the body_html descriptions. They are closer to full AI visibility than any other brand audited.

The brand

UpCircle Beauty was founded in 2016 by siblings Anna and William Brightman. The brand's core proposition is transforming by-products from the food and drinks industry into natural skincare. They started with used coffee grounds and have since expanded to upcycle 50+ leftover ingredients from the argan, tea, juice, coffee, date, apricot, olive, and wood industries. More than 450 tonnes of would-be waste ingredients repurposed.

B Corp certified since November 2022 with a score of 107.2 (median is 50.9). Plastic negative certified, 95-99% plastic-free packaging. All products UK-made, vegan, cruelty-free. Crowdfunded EUR 480,000 via Crowdcube in 2022 to finance US expansion. Now stocked at Ulta Beauty (768 doors), Holland & Barrett, Boots, Sainsbury's, Credo Beauty, Whole Foods, and Amazon. Trustpilot 4.6/5 with 2,354 reviews.

I selected UpCircle because it operates at the intersection of sustainability narrative and product efficacy — the exact tension this audit programme is designed to test. Does a strong brand story translate into AI visibility? Or do AI agents only care about product data?

The answer, it turns out, is both.

The test

I ran 5 queries across ChatGPT, Gemini, and Copilot. 15 tests total.

The queries:

1. "What's a good sustainable face serum with natural ingredients?"
2. "Can you recommend an eco-friendly body scrub?"
3. "What's a good face oil with rosehip for dry skin?"
4. "I want a moisturiser from a zero-waste beauty brand. Suggestions?"
5. "Can you recommend a coffee-based face scrub that's gentle enough for daily use?"

The results

UpCircle surfaced on 8 out of 15 tests. More than every other brand in this audit programme combined.

Query 1 (sustainable face serum): 2 out of 3. Gemini placed UpCircle first. Copilot placed it first in its "additional well-regarded sustainable serums" category. ChatGPT recommended Satin Naturel, Gaia Vegan, Natural Spa, and Conscious Skincare — brands with a fraction of UpCircle's distribution.

Query 2 (eco-friendly body scrub): 1 out of 3. Gemini placed UpCircle first, alongside AKT, Omaly, and Herbivore Botanicals. ChatGPT recommended Odylique, Outback, Organic Shop, and WiDEYE. Copilot recommended NEOM, Dr. Botanicals, The Body Shop, and Tree Hut.

Query 3 (rosehip face oil for dry skin): 0 out of 3. This was the control query — a product-attribute query with no sustainability signal. ChatGPT recommended Trilogy, The Ordinary, and Dr. Organic. Gemini recommended The Ordinary, Pai Skincare, and Trilogy. Copilot recommended Pai Skincare, Grown Alchemist, Mario Badescu, and Evolve. UpCircle does not sell a dedicated rosehip face oil, so this result is partly expected. But it also shows the limits of brand-level sustainability signals. The moment the query dropped sustainability language and asked purely about product attributes, UpCircle disappeared entirely.

Query 4 (zero-waste moisturiser): 3 out of 3. The only perfect score in this audit programme. ChatGPT placed UpCircle first. Gemini placed it second in its "Circular Champions" category. Copilot placed it first. This query perfectly matched UpCircle's brand identity: zero-waste, sustainable, moisturiser. All three platforms agreed.

Query 5 (coffee-based face scrub): 2 out of 3. Gemini placed UpCircle first, ahead of OGX and Frank Body. Copilot placed it first, with a shopping tile. ChatGPT recommended Dr Botanicals, Organic Shop, Nature Spell, and Biotique — missing the brand that literally started with upcycled coffee grounds.

Platform breakdown

The platform-level pattern is striking:

• Gemini: 4/5 queries, all in first position. The strongest platform by far. Gemini appears to weight sustainability credentials and niche editorial coverage heavily — the same pattern from the fashion audits, where Gemini consistently favoured brands with third-party sustainability verification (Good On You, Fair Wear, B Corp).
• Copilot: 3/5 queries. Appeared in both recommendations and shopping tiles. First position on two queries.
• ChatGPT: 1/5 queries. But it was a significant result — UpCircle is the first beauty brand in this entire audit set to surface on ChatGPT. The zero-waste query (Q4) broke through where Summer Fridays, Dr Sam's, and Herbivore could not.

The two-layer pattern

UpCircle's results reveal a pattern that no other audit has shown this clearly.

Layer 1: Sustainability identity queries — UpCircle dominates. When the query contains words like "sustainable," "eco-friendly," "zero-waste," or "coffee-based" (which maps to UpCircle's founding story), AI agents find the brand and often place it first. This is where editorial presence in sustainability publications (CNN Underscored, CleanHub, B Corp directory), the B Corp certification, the Trustpilot 4.6/5 with 2,354 reviews, and the upcycled ingredient narrative all combine to create strong signals.

Layer 2: Product attribute queries — UpCircle vanishes. When the query drops sustainability language and asks about product attributes ("rosehip face oil for dry skin"), UpCircle is invisible. Pai Skincare, Trilogy, and The Ordinary appeared instead. These brands have richer product-level data: full INCI lists in crawlable descriptions, skin type suitability, usage instructions, concentration percentages. The product data layer, not the brand narrative layer, determines visibility for attribute queries.

This is the first audit where I can see both layers working. Dr Sam's and Herbivore scored 0/15 — they failed on both layers. Summer Fridays scored 2/15 — thin product data with moderate editorial presence. UpCircle scores 8/15 because it has saturated Layer 1. It loses the remaining 7 because Layer 2 is thin.

Why Layer 1 works for UpCircle

Three things combine:

1. The B Corp and sustainability story is everywhere AI agents look. B Corp directory listing. CNN Underscored recommendation. CleanHub feature. Eco Skincare UK feature. The upcycled ingredient story is told consistently across the brand's own site, retail partners, press, and certification bodies. When an AI agent processes "sustainable skincare brand," it finds UpCircle cited by authoritative, third-party sources.

2. Trustpilot 4.6/5 with 2,354 reviews is a strong external signal. The second-strongest Trustpilot in the beauty audit set. This is a claimed profile with 88% five-star ratings. Compare: Summer Fridays 1.8/5 (30 reviews, unclaimed), Herbivore 3.9/5 (8 reviews), Dr Sam's 4.6/5 (641 reviews). UpCircle's Trustpilot volume is 4x Dr Sam's and the only one in the beauty set with enough reviews to be a meaningful signal. Trustpilot is one of the few review signals that exists outside a brand's own site and outside major retailers. AI agents appear to weight it.

3. The product tags are excellent. UpCircle's Shopify tags include Sensitive Skin, Acne-Prone, Collagen Banking, Glass Skin, Double-Cleansing, Skin Barrier Repair, and dozens more. These map directly to how consumers search. While tags are not visible in structured data, they power collection pages and internal search — which crawlers can navigate.

Why Layer 2 fails

1. No INCI ingredient lists in body_html. The Peptide Serum has 18 ingredients. The Cleansing Balm has 19. The Body Scrub has 15. None of these appear in the Shopify body_html field. An AI agent asked "does this contain niacinamide?" cannot answer from the crawlable data, even though the Peptide Serum does contain niacinamide and lists it in the product tags.

2. No usage instructions in body_html. None of the three products tested include how-to-use guidance in the description. "How do I use a peptide serum?" is a common AI query pattern. Even brief instructions ("Apply 2-3 drops to clean skin morning and night before moisturiser") would add queryable content.

3. Thin descriptions on key products. The Coffee Body Scrub — the original product that launched the brand, the product most associated with UpCircle's identity — has only 60 words of body_html. At GBP 18.99, this is a hero product with a 450-tonne upcycling story behind it. The description does not mention how often to use it, what skin types it suits, or how it differs from the peppermint variant.

4. Product comparison context is absent. UpCircle sells two body scrubs (lemongrass vs peppermint), two cleansers (balm vs milk), and multiple serums. The descriptions do not explain the differences. An AI agent asked "which UpCircle scrub should I use?" cannot answer this.

What UpCircle could do, in priority order

Phase 1 (1-2 days):

• Add INCI ingredient lists to all body_html descriptions. The data exists (verified via INCIDecoder and Amazon listings). It needs to be in the Shopify description field. This single change would make UpCircle competitive on ingredient-specific queries.
• Add usage instructions to all body_html descriptions. Brief, routine-contextual instructions ("Use morning and night after cleansing") would address a common query pattern that the current descriptions completely miss.
• Enrich the Coffee Body Scrub description. 60 words for the brand's founding product is too thin. Add the upcycling story, usage frequency, skin type suitability, and differentiation from the peppermint variant.

Phase 2 (3-5 days):

• Verify and enrich JSON-LD structured data. This audit could not confirm whether Product JSON-LD is present (JS-rendering limitation). If it is, extend it with ingredient and skin concern attributes. If it is not, implement it — Shopify's default is easy to activate.
• Surface the tag taxonomy in visible page content or structured data. Tags like Sensitive Skin, Acne-Prone, and Glass Skin are excellent but invisible to AI agents reading page-level data.
• Configure the on-site review system to output aggregateRating in JSON-LD. The brand claims 25,000+ five-star reviews, but no review platform was detected in server-rendered HTML.

Phase 3 (2-3 weeks):

• Add product comparison content for similar products. Two body scrubs, two cleansers, multiple serums — customers and AI agents need help choosing.
• Target mainstream beauty editorial roundups. UpCircle's editorial presence is strong in sustainability publications but weak in Vogue, Glamour, Allure, and Byrdie. Mainstream beauty roundups are a primary source for AI agents answering generic product queries.
• Leverage Amazon review signals. Amazon listings show "200+ bought" and "300+ bought" indicators. Ensure Amazon descriptions match enriched Shopify descriptions.

The competitor landscape

The sustainable skincare AI space has its own competitive set, distinct from mainstream skincare (CeraVe, La Roche-Posay) and prestige clean beauty (Drunk Elephant, Tatcha).

Pai Skincare appeared on 3 out of 5 queries across Gemini and Copilot. Another UK sustainable skincare brand with strong editorial presence in clean beauty roundups. Emerging as UpCircle's closest competitor in AI visibility for natural skincare queries.

Evolve Organic Beauty appeared on multiple queries. A small UK brand likely benefiting from the same sustainability editorial signals as UpCircle.

Biossance appeared on 2 queries across Copilot and Gemini. US-based, Amyris-owned, squalane-focused. Stronger mainstream editorial presence than UpCircle.

Ethique appeared on the zero-waste query alongside UpCircle. New Zealand-based solid beauty bar brand. Direct competitor in the zero-waste space.

The Ordinary appeared on Q3 (rosehip face oil) across all platforms. Their ingredient-forward naming convention continues to give them visibility on ingredient-specific queries across every category tested.

The pattern: UpCircle, Pai Skincare, Evolve, and Biossance are emerging as the AI-recommended brands for sustainability-filtered queries. UpCircle leads on Gemini and zero-waste queries. Pai leads on natural ingredient queries.

The pattern holds across categories

UpCircle is the beauty equivalent of Neem London.

In the fashion audits, Neem London — a small UK sustainable menswear brand — outperformed brands with 10x its revenue and recognition on AI visibility. Nudie Jeans outperformed Hiut Denim. ASKET outperformed Taylor Stitch. In beauty, UpCircle outperforms everyone.

The common thread across 10+ brand audits in two categories: B Corp or equivalent third-party sustainability verification, strong Trustpilot, niche editorial presence in sustainability publications, and a differentiated sustainability story. These signals consistently produce AI visibility regardless of brand size, revenue, or social media following. This is now a confirmed pattern, not an anomaly.

Close

UpCircle Beauty scored 8 out of 15 on AI shopping queries. The highest in this audit programme. Higher than Summer Fridays (2/15) with its Allure awards and 1.1 million followers. Higher than Dr Sam's (0/15) with its Harley Street dermatologist founder and JSON-LD structured data. Higher than Herbivore (0/15) with its $50M revenue and Sephora distribution.

A B Corp-certified upcycled skincare brand from the UK, with ~280K social media followers and crowdfunded expansion capital, outperformed every prestige beauty brand on AI visibility.

But the 8/15 tells a precise story. UpCircle wins when the query contains sustainability language. It vanishes when the query is purely about product attributes. "Zero-waste moisturiser" — first on all three platforms. "Rosehip face oil for dry skin" — invisible on all three.

The sustainability narrative has been converted into AI visibility. The product data has not. UpCircle is the closest any brand has come to getting this right. The gap between 8/15 and 15/15 is not brand building, editorial coverage, or certification. It is INCI lists, usage instructions, and skin concern context in the Shopify description field.

The brand story opens the door. The product data keeps it open.