Five categories.
Five different AI marketplaces.

One brand owns the AI recommendation layer across nearly every market.

Across 8,324 AI conversations in May 2026, run across all six leading models and 38 U.S. geographic panels, Club Pilates is the dominant brand AI recommends: top-mentioned in 36 of 38 geographies (94.7%), in every one of the six models, at 57.6% recommendation share. More than half of every AI conversation about Pilates studios surfaces Club Pilates by name.

What the data measured

• 8,324 distinct AI responses to Pilates recommendation prompts (canonical May 2026 panel)
• Six leading models: ChatGPT, Claude, Gemini, Grok, Perplexity, Microsoft Copilot
• 38 user-location panels: corporate HQ markets, mid-tier metros, suburban submarkets, franchise-density geographies
• March and April included as trajectory context only (partial panel); May is canonical for all leadership claims

Top dedicated Pilates brands (May 2026)

The structure is barbell-shaped. Club Pilates at 57.6%, the challenger at 16.2%, then a cliff to a tight second tier between 2% and 5%. There is no middle.

Substitute and adjacent operators

AI does not limit Pilates recommendations to dedicated Pilates brands. When it cannot fill the top of the list with specialty brands for a market, it reaches for adjacent and substitute operators.

Adjacent and substitute brands collectively earn 18.9% of the May surface, more than the entire dedicated-Pilates second tier (ranks 3 through 8) combined. When AI cannot name a third or fourth dedicated Pilates brand for a market, it expands the answer to fitness operators, a competitive risk dedicated brands rarely address.

Month-over-month movement

The April-to-May jump is panel-driven, not market-driven: March and April were partial-panel (1 to 3 models, 12 to 18 geos), May is the full panel (6 models, 38 geos). The signal is consistency, not improvement. The brands that gain as the panel widens are those with the broadest cross-model, cross-geo presence. Club Pilates gains 22 points; the challenger gains 11. Every other brand is flat or undetectable in partial months. That trajectory is itself a measurement: the challenger is the only #2 candidate whose presence holds up as the surface widens.

Geographic concentration

Club Pilates is #1 in 36 of 38 geographies. The two exceptions: one suburban Texas market where the challenger wins, and Boston, MA, where Boston Body Pilates, a local independent, holds the top spot. No substitute fitness operator wins #1 anywhere. Club Pilates is #1 in every Tier 1 metro, every suburban submarket, and every franchise-density market measured. Geographic dominance is effectively universal.

Model-by-model consistency

Club Pilates is #1 in all six models with no disagreement about the leader. The challenger holds #2 in four models and #3 in Copilot. Claude is the exception, favoring adjacent brands (Pure Barre, CorePower) in its #2 and #3 slots. Copilot is the only model where a substitute (YMCA) outranks any dedicated Pilates brand other than the leader.

Strategic implication

Nationally, the challenger is the structural #2 dedicated Pilates brand in 5 of 6 models: 16.24% share versus the next dedicated brand (Solidcore) at 5.05%, a 3.2x lead over the rest of the field, and the only #2 candidate whose share expanded as the panel widened.

Locally, the story inverts. Across 12 measured markets where the challenger has known physical presence, Club Pilates is #1 in 11 of them. The challenger wins #1 in exactly one of its own markets. In the challenger's own corporate-headquarters market, Club Pilates earns 90 mentions as the top recommendation. The challenger does not place #1 at home.

The implication is precise: a brand can earn national #2 recognition while losing the "recommended first" position in 11 of 12 of its own markets, including headquarters. The gap is not category brand-awareness. It is a local-AI-presence gap that gets measured every time a prospect asks ChatGPT, Claude, Gemini, Grok, Perplexity, or Copilot for a Pilates studio near them. Closing it runs through identifying the citation surfaces and content patterns that put the leader in the top position per market, then executing that work market by market over 90 to 180 days.

The most contestable category measured. Local presence is everything.

Across 2,516 AI conversations in April 2026, run across four models and 18 panels, no brand holds a dominant position. The leader, SkinSpirit, earns 3.4% share. Every one of the 18 markets surfaces a different local champion as its #1 med spa. No brand wins #1 in more than one market. This is the inverse of Pilates: there is no national AI default.

What the data measured

• 2,516 distinct AI responses to Med Spa recommendation prompts (canonical April 2026 panel)
• Four of six leading models: ChatGPT, Claude, Gemini, Perplexity
• 18 user-location panels: Tier 1 metros and suburban submarkets
• Methodology note: ChatGPT's top-2 raw results and Claude's #2 raw result were treatment-type or generic extraction artifacts, excluded from analysis. The dedicated competitive set is intact with artifacts removed

Top dedicated Med Spa brands (April 2026, artifacts excluded)

The combined top 10 hold roughly 21% of the April surface, less than Massage Envy alone in Massage, and less than half of what Club Pilates alone holds in Pilates. The category top is barely consolidated at all.

Substitute and adjacent operators

Here the adjacent layer is dermatology practices and aesthetic surgery offices. AI recommends Westlake Dermatology, Houston Dermatology Associates, Buckhead Facial Plastic Surgery, La Fontaine Aesthetics, and similar clinical operators for injectables, lasers, and aesthetic services. They earn roughly 3% of the April surface, small in absolute terms but meaningful given the fragmentation.

Geographic concentration

The most fragmented per-geo distribution in the report. Eighteen markets, eighteen unique champions, zero overlap.

No brand recurs as #1 in any other market. This is a structural finding, not an artifact: Med Spas has no national AI default at the recommendation layer.

Strategic implication

This is the most contestable category measured, and its implication runs exactly opposite to Pilates. The national #1 holds 3.4%, so there is no dominant brand to displace. Every market goes to a different local operator. The four models disagree on the "true #1": SkinSpirit, Skin Laundry, and SkinLogic each hold #1 in at least one model, but none is consensus. Adjacent dermatology and aesthetic-surgery practices earn ~3%.

The implication for brands is the most actionable in the report. Market-level AI dominance is achievable for any single-location or small-chain med spa willing to invest in citation surface, content authority, and local third-party validation at home. There is no Club Pilates, no Massage Envy, not even an Elements to compete against. A 50-mention concentration in a single market can win the local #1 spot, as Cienega does in Los Angeles, Yofi in Philadelphia, and Persona in Houston. A 75-mention concentration across two markets would make a brand the only multi-market #1 winner in the category. National distribution adds expense without an entrenched default to fight, because none exists.

AI defers to institutional medical authority, not commercial brand authority.

Across 2,272 AI conversations in May 2026, run across five models and 18 panels, the recommendation layer behaves unlike any other category here. The leading brand is not a commercial practice. It is a hospital system. Of the top 20 most-recommended entries, 14 are hospital systems, academic medical centers, or insurance companies. Only one dedicated private practice, Health Loft at 4.4%, breaks the top 5.

What the data measured

• 2,272 distinct AI responses to Nutritionist and Dietitian recommendation prompts (canonical May 2026 panel)
• Five of six leading models: ChatGPT, Claude, Gemini, Grok, Perplexity
• 18 U.S. panels: Tier 1 metros and select strategic submarkets
• Because this category produced more institutional and non-commercial entities, additional entity normalization was applied before analysis

Top dedicated nutrition brands

The dedicated private-practice layer is sparse compared to every other category.

Health Loft is the only dedicated multi-location practice with material national share, roughly 4x the next dedicated brand. Beyond it, the private-practice layer is a long tail of single-location practices in the 0.3% to 1% range, surfacing in 1 to 3 models inconsistently.

The dominant layer: hospital systems and academic medical centers

In every other category, dedicated specialty brands hold the top and substitutes take a secondary share. Here the structure is inverted. Hospital systems, academic medical centers, and insurance companies are the dominant layer.

The institutional layer collectively earns more than 25% of the surface, dwarfing the private-practice layer. Fourteen of the top 20 entries are institutional, and most surface in 4 of 5 or all 5 models. AI treats institutional medical authority as the trusted pathway for dietitian and nutritionist searches in a way it does not in any other category. Insurance companies (Aetna, Blue Cross Blue Shield, UnitedHealthcare, Cigna) appear because AI recommends consumers contact their insurer for covered dietitian services rather than naming a practice, a behavior unique to this category.

Strategic implication

For dedicated private practices, the path to share runs through breaking the institutional default. AI's first answers surface hospitals, academic centers, and insurance pathways; a practice must displace some of that. Health Loft proves it is possible at 4.4% across 4 of 5 models. For single-location practices the per-market opportunity is genuinely open, and no local private practice has yet established cross-model consensus in its home market. The shape mirrors Med Spas: market-level dominance is achievable through local citation surface and content authority.

For hospital systems and academic medical centers, this is an unusually durable moat: high cross-model consensus, institutional recognition translating directly into AI default, minimal commercial challenge. The posture is defensive: monitor for emerging private-practice brands gaining cross-model coverage, and keep investing in the structured-data and citation patterns that produced the default.

Of the five categories, this is the one where commercial brand-building produces the least direct AI recommendation lift. Private practices face a structurally harder surface than they would in any other measured category.

What this changes for the people deciding category strategy.

The structure governing a category determines what posture works in it. These are organized by audience, because the people deciding strategy think in roles, not frameworks.

For CEOs

• AI recommendations are becoming a category-level competitive surface. The AI's first answer for "best [category] near me" is increasingly the buyer's first impression, formed before any website visit. These categories show AI defaults already exist, are durable, and produce winners and losers.
• National brand strength does not automatically convert into per-market AI default. The Joint holds 23% national share but wins only 25% of markets. That national-to-local conversion gap is invisible in conventional marketing reports and material at the CEO level.
• AI-layer consolidation patterns can foreshadow commercial-layer consolidation. Default-Dominant Markets signal late-stage consolidation; Local Fragmentation Markets signal early-stage opportunity.

For CMOs

• Search visibility and AI recommendation share are no longer the same measurement. A brand can rank in organic search and be invisible in AI answers, or hold strong presence in some models and be absent in others. ARDI measures the second surface; legacy SEO and AI-visibility tools do not.
• Investment that produces share in a Default-Dominant Market does not produce share in a Local Fragmentation Market. National spend compounds into AI default in some structures and produces no measurable lift in others. Market Structure belongs in the media-mix decision, not as an afterthought.
• Cross-model coverage is a separate axis from cross-channel coverage. A brand can be consistent in ChatGPT and Perplexity and absent in Claude. Cross-model consensus is one of the strongest predictors of durable AI position.

For Franchise Operators

• Franchisor-level brand strength does not produce franchisee-level AI default. The Joint loses local #1 in 18 of 24 markets despite physical clinics in many of them. The "national franchise" advantage does not transfer to local AI recommendation by default.
• Converting a location from local-independent-#1 to brand-#1 is local citation surface, local content authority, and local third-party validation. The investment is location-by-location, the measurement monthly, the conversion observable per geography.
• Franchise competitive analysis should include per-market AI recommendation share alongside unit economics. A location losing local AI #1 to a single-location independent is a leading indicator of acquisition friction that never appears in same-store sales.

For Multi-location Brands

• Oligopoly and National-Recognition-with-Local-Leakage markets are a structurally different surface than Default-Dominant markets. Multi-location strategy should be calibrated to the structure that governs the brand's primary category.
• In National-Recognition-with-Local-Leakage markets, the gap between national share and per-market #1 conversion is the most leveraged improvement available. Closing it city by city compounds faster than national share growth and is more defensible.
• Regional concentration is a viable distribution strategy in Oligopoly markets. Elements Massage proves it, winning three suburban Texas markets and Phoenix without a single Tier 1 metro.

For Local Operators

• Local AI default is achievable in Local Fragmentation markets and in 75% of Chiropractor markets without national distribution. The surface is open in three of every four Chiropractor markets and in every Med Spa market measured.
• The investment threshold to win local AI default in a fragmented or contested market is materially lower than to displace a national leader in a Default-Dominant market. Single-location operators can earn per-market #1 with focused local citation, content, and validation work.
• Measuring per-market AI recommendation share is the first move. Without it, a local operator cannot tell whether it is already winning or already losing share to a competitor that has done the work.

For Healthcare & Regulated Categories

• Institutional Authority markets behave differently from commercial ones. AI's deference to hospitals, academic centers, and insurance pathways is structural, not incidental. Commercial specialty brands in regulated categories should expect the default to favor institutions unless they build equivalent institutional-grade citation surface.
• Health Loft shows commercial brands can break the institutional default: 4.4% national share across 4 of 5 models, the only multi-location private nutrition practice with meaningful AI presence. The pattern that produced this exception is worth studying.
• For regulated-category brands weighing AI-recommendation investment, the work should be informed by which structure governs the category and what institutional-grade citation surface the incumbents already hold.