How AI Works Service Research ↳ Protein Powder ↳ Mattress About Get in Touch
Category: Supplements AI Surface: ChatGPT Date: Feb 9, 2026 Query set: Broad + constrained queries Variable tested: Query type & constraint specificity

How AI Selects
"Best" Protein Powder Brands

A controlled experiment mapping how ChatGPT selects and frames protein powder brands across high-intent recommendation queries. The goal was not to evaluate product quality — but to identify the structural patterns AI actually rewards.

Core Finding
In one sentence

In broad "best protein powder" queries, ChatGPT consistently favours brands that function as stable category anchors — those widely framed as safe, industry-standard, and general-purpose — while excluding highly specialized or narrowly positioned brands unless the query explicitly requires them.

If a brand is not clearly positioned as either a consensus-backed category anchor or an unmistakable fit for a specific constraint (e.g., clean label, no artificial sweeteners), it may be ineligible for inclusion in AI-generated recommendations — regardless of product quality. Without deliberate signal clarity, brands risk invisibility at the moment of decision, where purchase intent is highest.

Experiment Design

How we ran this experiment

Five "best" queries were run within a single ChatGPT session to reduce contextual variability. No search engine data or SEO rankings were included. The focus was solely on AI-generated responses.

01

Query Set

What is the best protein powder? · For muscle gain · For women · Without artificial sweeteners · Best brand

02

Environment

Single ChatGPT session. Controlled context. No prior search history or brand preferences introduced.

03

Objective

Identify structural patterns in brand selection — not evaluate product quality or rank brands by merit.

Brand Frequency

Which brands AI kept returning to

Across 5 queries, brand appearance was highly uneven. Frequency indicates structural dominance — the brands AI defaults to when forced to choose.

Brand appearances across 5 queries
Optimum Nutrition
4/5
4
PEScience Select
3/5
3
Myprotein Impact Whey
2
Naked Whey
2
Allmax
2
MuscleTech Nitro Tech
1
Isopure Zero Carb
1
LeanFit Whey
1
Which brands appeared in which queries
Brand Best Overall Muscle Gain Women No Sweeteners Best Brand
Optimum Nutrition
PEScience Select
Myprotein Impact Whey
Naked Whey
Allmax
MuscleTech Nitro Tech
Pattern Analysis

What we found

These are not tactics — they are signals AI appears to reward. Each pattern was identified from direct observation, not assumption.

Pattern 01

Broad queries favour consensus anchors

In unconstrained "best" queries, ChatGPT consistently surfaced brands framed as stable, widely recognized, and general-purpose. Optimum Nutrition appeared in 4 of 5 queries — described as "classic," "industry standard," and "widely recommended." These brands function as category anchors, not niche specialists.

Framing language observed: "Classic best-selling" · "All-purpose option" · "Widely recommended" · "Industry standard"

Pattern 02

Explicit constraints override anchor status

When an ingredient constraint was introduced ("without artificial sweeteners"), anchor brands were excluded entirely — including Optimum Nutrition, which dominated every other query. Only brands with unmistakable clean-label identity qualified. Anchor status does not override absence of constraint-specific signals.

Optimum Nutrition (4/5 overall) did not appear. Constraint winners: Naked Whey, Allmax IsoNatural, Isopure Zero Carb, LeanFit — all with explicit "no artificial sweeteners" associations.

Pattern 03

Specialization narrows contextual eligibility

Highly engineered or performance-enhanced formulations appeared only in niche queries (e.g., muscle gain). They did not surface in broad "best" contexts. Specialization may increase relevance in narrow categories while reducing surface area across general queries.

MuscleTech Nitro Tech (creatine-enhanced) appeared only in muscle gain. Allmax Isoflex (fast-absorbing isolate) appeared in muscle gain and clean contexts only. Neither appeared in broad "best" queries.

Structural Model

How AI selection logic works

Across all patterns, AI brand selection appears to operate under layered logic. Brand visibility is conditional, not absolute.

Rule 01

Broad queries → consensus anchors

AI defaults to widely trusted, general-purpose brands when no constraint is present.

Rule 02

Constrained queries → eligibility filters

Explicit constraints apply hard filters. Anchor status does not compensate for lack of alignment.

Rule 03

Specialization → narrow eligibility

Specialized positioning increases depth in niche queries but reduces surface area in general ones.

What This Means for Brands

Signal gaps that determine visibility

Category anchor framing

Winning signal

Widely reinforced "industry standard" positioning. Framed as safe, general-purpose, and all-around. Described with consensus language across multiple sources.

Constraint alignment

Winning signal

Explicit and repeated association with a specific constraint (e.g., "no artificial sweeteners," "clean label"). Ambiguity results in exclusion when constraints are applied.

Role clarity

Missing signal

Brands trying to simultaneously signal performance, value, and premium authority lack a stable, repeatable role — reducing summarizability in AI responses.

Notable absences

Missing entirely

Ghost Protein, Transparent Labs, Orgain, Garden of Life, and Vega did not appear in any query — despite category relevance and strong consumer awareness.

Closing Observation
The takeaway

Across the protein powder category, AI does not optimise for the most advanced formulation, the cleanest ingredients, or the lowest price. Instead, it favours brands that are easy to summarise, widely reinforced, and safely aligned with the context of the query. Visibility appears to be a function of signal clarity rather than brand familiarity alone.

Run Your Own Experiment

Find out where your brand stands.

Every category behaves differently. We'll run a controlled experiment on your brand and tell you exactly what AI sees — and what it doesn't.