Stat Strip

• AI platforms analyzed: 6 major LLM ecosystems
• High-intent prompt clusters: 20+
• Observations analyzed: 20,000 modeled prompt interactions
• Category focus: Clean makeup, skincare-infused cosmetics, sensitive-skin beauty, and hybrid beauty brands

Answer Capsule

AI recommendation power in clean makeup appears to be concentrating around a small set of brands that combine strong product-specific authority, editorial reinforcement, skincare framing, and broad cross-platform familiarity. Rare Beauty, Kosas, ILIA, Tower 28, and e.l.f. Cosmetics consistently appear in high-intent recommendation environments, while legacy “clean beauty” positioning alone no longer guarantees shortlist inclusion.

Executive Summary

The clean makeup category is undergoing a structural shift inside AI-assisted discovery environments.

Historically, beauty brands competed through retail placement, influencer visibility, editorial coverage, and social virality. But large language models are changing how product consideration forms. Increasingly, buyers are not starting with TikTok or Google. They are asking ChatGPT, Gemini, Copilot, or Perplexity direct purchase-intent questions:

• “What’s the best foundation for acne-prone skin?”
• “Best concealer for mature skin?”
• “What clean beauty brands are actually worth it?”
• “Which makeup brands are best for sensitive skin?”

The brands that repeatedly get advanced into those recommendation shortlists are beginning to accumulate disproportionate AI-era visibility advantages.

The strongest category signal is not who appears most often. It is who gets recommended in buyer-choice moments.

Current directional evidence suggests that recommendation power in clean makeup is concentrating around a relatively small set of hybrid brands:

• skincare-infused makeup brands,
• dermatologist-compatible positioning,
• ingredient-conscious formulations,
• “clean but effective” framing,
• and brands reinforced by editorial and community validation.

Rare Beauty, Kosas, ILIA, Tower 28, and e.l.f. Cosmetics repeatedly surface across multiple commercial-intent prompt clusters. Meanwhile, many traditional prestige makeup brands still appear frequently but are often framed as secondary options, alternatives, or specialist picks rather than default recommendations.

This matters because AI-assisted shortlist formation compresses consideration sets. If a brand is absent — or merely “present but not prioritized” — it may lose consideration before the buyer ever reaches Sephora, TikTok, YouTube, or Google.

The AI Discovery Shift in Clean Makeup

The clean makeup market was already crowded before AI recommendation systems entered the discovery layer.

Now the category faces a new dynamic:
AI systems are collapsing hundreds of possible beauty choices into a shortlist of 3–8 brands.

That changes the economics of visibility.

In traditional search, a brand could survive with:

• strong SEO,
• retailer placement,
• influencer activity,
• or paid media.

But AI systems increasingly synthesize:

• editorial reviews,
• retailer signals,
• dermatologist narratives,
• ingredient framing,
• product awards,
• Reddit/community sentiment,
• and comparison-style content.

That synthesis layer creates recommendation concentration.

A brand can still be highly visible online and still fail to become a preferred AI recommendation candidate.

The clean beauty category is especially exposed to this because many buyer prompts involve:

• trust,
• safety,
• skin sensitivity,
• ingredient quality,
• acne compatibility,
• mature skin,
• fragrance concerns,
• and “makeup + skincare” hybrids.

Those are exactly the kinds of nuanced recommendation environments where AI systems rely heavily on external evidence architecture.

Directional Category Leaders

Current recommendation patterns suggest several brands are outperforming the broader category in AI-assisted buying environments.

Rare Beauty

Framing: Leader

Rare Beauty appears unusually resilient across multiple product categories:

• blush,
• highlighter,
• lip oils,
• under-eye brighteners,
• brow products,
• complexion products.

The brand benefits from broad familiarity plus strong “natural-looking,” “lightweight,” and “easy to wear” framing. That positioning aligns extremely well with conversational recommendation systems.

Rare Beauty also appears to benefit from strong editorial reinforcement and exceptionally high consumer familiarity.

Kosas

Framing: Strong option

Kosas performs especially well in prompts tied to:

• mature skin,
• skincare-infused makeup,
• hydration,
• lightweight complexion products,
• sensitive-skin compatibility.

The brand repeatedly surfaces in concealer, corrector, and brow-related recommendation environments.

Its strongest AI-era advantage may be semantic positioning:
“makeup + skincare hybrid.”

That framing is highly retrievable inside LLM recommendation systems.

ILIA

Framing: Strong option / specialist leader

ILIA appears strongest in:

• clean complexion prompts,
• sensitive skin prompts,
• skincare-first makeup environments,
• natural-finish makeup,
• brow products.

Notably, ILIA performs well when prompts contain qualifiers like:

• “clean”
• “non-comedogenic”
• “sensitive skin”
• “natural finish”
• “serum foundation”

That suggests the brand’s recommendation power is closely tied to trust-oriented and ingredient-aware buyer intent rather than broad mass-market dominance.

ILIA’s strongest signal is not sheer mention volume. It is contextual relevance inside high-trust buying moments.

Tower 28

Framing: Specialist option

Tower 28 appears repeatedly in:

• acne-safe prompts,
• sensitive skin prompts,
• serum concealers,
• lightweight complexion recommendations,
• cream bronzer discovery.

The brand’s recommendation footprint appears tightly connected to safety-oriented beauty discovery.

That positioning may become increasingly valuable as AI systems continue prioritizing “safe recommendation” logic in skin-related prompts.

e.l.f. Cosmetics

Framing: Strong option / value disruptor

e.l.f. occupies a unique role in the category.

The brand repeatedly appears not only in budget beauty prompts, but increasingly alongside prestige and clean-adjacent brands in:

• primers,
• lip products,
• brow products,
• eye brighteners,
• complexion recommendations.

That matters because AI systems often optimize for:

• accessibility,
• value,
• broad approval,
• and availability.

e.l.f.’s growing recommendation presence suggests affordable brands can still dominate AI discovery if they achieve strong cross-source validation.

The Buying Moments That Now Decide the Category

The clean makeup category is increasingly being decided inside a handful of high-intent prompt environments.

The highest-pressure clusters currently appear to include:

1. Sensitive Skin + Acne Compatibility

This is one of the strongest recommendation environments in the category.

Examples:

• “Best foundation for acne-prone skin”
• “Best makeup for sensitive skin”
• “Non-comedogenic makeup brands”

Brands repeatedly surfacing here include:

• ILIA,
• Tower 28,
• Kosas,
• Clinique,
• dermatology-adjacent products.

These prompts matter because they combine:

• trust,
• product performance,
• ingredient safety,
• and purchase intent.

2. Makeup + Skincare Hybrid Prompts

This cluster appears to be growing rapidly.

Examples:

• “Makeup with skincare benefits”
• “Hydrating concealer”
• “Serum foundation”
• “Clean makeup with SPF”

Brands benefiting most:

• ILIA,
• Kosas,
• Tower 28,
• Rare Beauty.

This may become one of the defining recommendation battlegrounds in beauty.

3. “Best Overall” Beauty Prompts

These are broad but commercially important:

• “Best beauty brand”
• “Best makeup brands”
• “Best clean makeup brand”

Here, recommendation concentration becomes more severe.

AI systems tend to favor:

• brands with broad familiarity,
• editorial reinforcement,
• strong retailer presence,
• and high cross-platform consensus.

Rare Beauty appears especially strong in this environment.

4. Mature Skin + Texture-Safe Makeup

This is another highly trust-sensitive cluster.

Examples:

• “Best concealer for fine lines”
• “Best under-eye corrector for mature skin”

Kosas repeatedly appears in these environments, often framed positively because of hydration-oriented and serum-like formulations.

5. Affordable Alternatives

Budget-conscious recommendation environments remain highly influential.

Examples:

• “Best affordable primer”
• “Drugstore clean makeup”
• “Best budget lip stain”

e.l.f. Cosmetics appears unusually resilient here.

That matters because recommendation systems frequently optimize for broad consumer accessibility rather than luxury positioning alone.

Why Recommendation Power Is Concentrating

Current evidence suggests recommendation concentration in clean makeup is being driven by citation architecture more than pure brand awareness.

The brands surfacing most often tend to have reinforcement across:

• beauty editorial ecosystems,
• retailer recommendation pages,
• dermatologist-adjacent content,
• creator ecosystems,
• product roundups,
• and comparison-driven beauty content.

Repeatedly cited environments include:

• Vogue,
• Allure,
• Marie Claire,
• Ulta,
• Sephora,
• Good Housekeeping,
• Forbes,
• and beauty review ecosystems.

Importantly, AI systems do not appear to rely solely on official brand websites.

Instead, recommendation power appears heavily shaped by:

• third-party validation,
• comparative product framing,
• and repeated cross-source consensus.

That creates a compounding effect.

A brand repeatedly endorsed across trusted beauty ecosystems becomes easier for AI systems to retrieve, compare, and confidently recommend.

The Category’s Most Visible Warning Sign

The strongest warning sign in clean makeup is that legacy “clean beauty” positioning alone no longer guarantees recommendation strength.

Several brands historically associated with clean beauty appear inconsistently across recommendation environments despite strong consumer awareness.

Meanwhile, hybrid brands with:

• clearer product utility,
• stronger editorial integration,
• broader demographic reach,
• and more conversational product framing

are increasingly capturing recommendation share.

This is especially visible in prompts involving:

• skin sensitivity,
• mature skin,
• acne compatibility,
• and “best overall” comparisons.

In AI-assisted discovery, a brand can still be visible and still be commercially absent from the shortlist.

That distinction may become one of the defining competitive pressures in beauty over the next several years.

What This Means for the Category

Several directional consequences are beginning to emerge.

AI Discovery Is Compressing Consideration Sets

Consumers may increasingly receive:

• 3–5 recommended brands,
• not 50.

That changes the economics of category competition.

Trust Architecture Matters More Than Brand Awareness

Editorial reinforcement, comparison ecosystems, dermatologist compatibility, and structured product explanations increasingly shape recommendation outcomes.

Product-Level Authority May Matter More Than Brand-Level Fame

Brands winning one high-intent category at a time:

• concealer,
• serum foundation,
• blush,
• lip oils,
• brow products

may accumulate broader recommendation authority over time.

“Safe Recommendation” Brands May Gain Structural Advantages

Brands associated with:

• sensitive skin,
• gentle formulations,
• dermatologist-safe positioning,
• non-comedogenic products

appear especially well-positioned for AI-era recommendation systems.

What This Public Benchmark Does Not Include

This public benchmark is directional and intentionally incomplete.

It does not include:

• full competitor threat profiles,
• exact recommendation share calculations,
• platform-specific visibility breakdowns,
• citation failure mapping,
• raw prompt inventories,
• recovery roadmaps,
• product-level ranking matrices,
• or proprietary weighting methodology.

The paid LLM Authority Index deep-dive includes:

• company-specific competitive diagnostics,
• AI recommendation gap analysis,
• citation architecture mapping,
• prompt-cluster opportunity modeling,
• and AI visibility recovery strategies.

Methodology and Disclaimers

This benchmark reflects a directional snapshot of AI-assisted discovery patterns within the clean makeup category during the May 2026 reporting window.

The analysis incorporates:

• modeled high-intent beauty prompts,
• recommendation environments,
• citation ecosystems,
• editorial co-occurrence patterns,
• and product recommendation observations across major AI systems.

Important limitations:

• This is not a definitive market census.
• Some recommendation environments have thinner platform coverage than others.
• Presence does not equal recommendation strength.
• Citation frequency does not equal endorsement.
• Economic significance estimates are directional rather than realized revenue figures.
• Competitive positioning is interpretive and based on observable recommendation patterns rather than platform-provided ranking APIs.

CTA

LLM Authority Index produces company-specific AI discovery diagnostics for brands evaluating how ChatGPT, Gemini, Perplexity, Copilot, and AI Overviews are reshaping category consideration.

The full enterprise report includes:

• competitive recommendation analysis,
• prompt-level visibility mapping,
• citation source diagnostics,
• AI shortlist displacement analysis,
• and recovery opportunity modeling.