Monthly LLM Visibility Benchmark — April 2026

What do 4 AI models cite in French beauty D2C?

April 23, 2026 15 brands × 4 LLMs 40 prompts × 3 runs How we score →

Top score — Typology (composite)

Brands below 50/100 — invisible to AI

−22

Avg pts drop for D2C brands on ingredient-specific queries

Visibility Heatmap

Scores by brand & platform

15 French beauty D2C brands ranked by average LLM visibility score across 4 AI models (0–100)

LLM Visibility Score — April 2026

ChatGPT

Perplexity

Gemini

Claude

Full Rankings

15 brands ranked — April 2026

Visibility scores reflect how often each brand is mentioned across ChatGPT, Perplexity, Gemini, and Claude in discovery, comparison, and purchase queries.

#	Brand	ChatGPT	Perplexity	Gemini	Claude	Overall	Trend

Key Findings

What the data reveals

An 11.8× gap separates 1st from last

Typology scores 94/100. Joëlle Ciocco scores 8/100. Same market, same vertical, same LLMs — an 11.8× visibility gap. Typology consistently appears across all 40 prompt categories. It has the broadest indexed content in both French and English — skincare actives, routines, ingredient guides, and editorial coverage — which correlates strongly with top-tier scores across all four models.

6 brands score below 35 — despite real community strength

Respire has 500k+ community members but scores 46/100. Seasonly has cult status but 33/100. Merci Handy, known for its vibrant brand, sits at 29/100. The gap isn't product quality — it's indexed content volume. A single tier-1 editorial feature or structured FAQ page can shift a score by 10–15 points.

Ingredient-specific queries expose a new gap

When users ask about retinol, vitamin C, hyaluronic acid, or niacinamide, the results diverge sharply. Brands with ingredient education content (Typology, Caudalie, Embryolisse) hold strong. D2C-only brands without ingredient guides drop 18–25 points on these queries vs. their general discovery scores.

Methodology — Read full methodology →

Analysis based on real LLM responses to 40 French consumer-style queries per brand — covering skincare discovery, ingredient-specific queries, purchase intent, brand comparison, and category exploration — across ChatGPT (gpt-4o-mini), Perplexity (sonar), Gemini (2.0 Flash), and Claude (3.5 Haiku). Scores are averaged across 3 independent runs conducted at 24-hour intervals for statistical stability. Visibility scored 0–100 based on mention frequency and position. Composite score weighted: ChatGPT 40%, Perplexity 25%, Gemini 20%, Claude 15%. Data reflects web-indexed content as of April 23, 2026. Rescanned: April 23–25, 2026. Next benchmark: May 2026.

15 brands 4 LLMs 40 prompts / brand 3 runs × 24h intervals FR prompts · Beauty vertical New methodology Scanned April 23, 2026

Free weekly intelligence

Get your brand's visibility score — every Monday

Stop guessing if AI mentions you. Benchfolk scans 4 LLMs — ChatGPT, Perplexity, Gemini & Claude — and sends you a score, a benchmark, and 3 things to fix. Just a 90-second read.

Get your free Scorecard — in 24h →

One free scan. No credit card. Upgrade to Intel for weekly tracking.