Research library · Published 7 September 2026

Can Emerging D2C Brands Become AI-Recommendable Before They Become Famous?

An Exploratory Study of ChatGPT Buyer-Intent Shortlists in India

By Sanchari Sarkar, Digital Platform 271

18 brands · 15 emerging + 3 controls · 3 prompts × 3 runs · 162 brand-prompt-run observations

Abstract

This exploratory study investigates whether emerging Indian D2C brands can enter ChatGPT buyer-intent recommendation shortlists before achieving broad mass-market recognition. Eighteen brands were examined, including 15 emerging brands and 3 control brands, across category-specific buyer-intent prompts. Each brand was tested through three frozen prompts repeated across three runs, producing nine recommendation observations per brand.

The study evaluated Recommendation Readiness (RR), External Evidence Score (EES), and Prompt Fit Score (PFS). Only 2 of the 15 emerging brands received any recommendation: Mezame appeared in 7 of 9 runs and Cinta Kids appeared in 4 of 9. RR and EES were positively associated with shortlist frequency, while an equally weighted standardized RR + EES composite showed a stronger rank association than either measure individually. Because the sample is small and EES and PFS were developed post-hoc, the findings are exploratory rather than causal.

Key findings

  • 2 of 15 emerging brands achieved any shortlist visibility.
  • Recommendation Readiness vs recommendation frequency: Spearman ρ = 0.698; exact permutation p = 0.0013.
  • External Evidence Score vs recommendation frequency: ρ = 0.665; exact permutation p = 0.0028.
  • Standardized equal-weight RR + EES vs recommendation frequency: ρ = 0.745; exact permutation p = 0.00035.
  • All observed shortlist entries occurred under maximum Prompt Fit in this sample, but 18 of 28 maximum-fit cases still produced no recommendation.

Brand-level outcomes

BrandGroupRREESRecommendations
BenticaEmerging510/9
Dreamy AtomsEmerging6820/9
MezameEmerging7087/9
Byora HomesEmerging4510/9
KiddiKindEmerging5930/9
Yellow NaturalsEmerging4520/9
AyethicEmerging5000/9
Roslyn by DemiEmerging5200/9
SitayyaEmerging5920/9
Slay SkincareEmerging1800/9
Beyond BasicEmerging2750/9
Wild DateEmerging5740/9
Cinta KidsEmerging7024/9
Casa Decor ShopEmerging5510/9
NuvieEmerging/funded5520/9
The FormulaRxControl5997/9
Nesta ToysControl10073/9
The Gourmet JarControl6851/9

Interpretive framework

Query Relevance → Owned Recommendation Evidence → External Corroboration → Competitive Shortlist Entry → Recommendation Stability.

The arrows represent a conceptual sequence for interpreting observed outcomes. They do not represent demonstrated causal pathways, a validated ranking model, or the internal architecture of ChatGPT.

Limitations

The study uses a small purposive cross-category sample, three prompts and three runs per brand/category. EES and PFS were developed post-hoc. Prompt-level observations are nested within brands. Broad-market recognition was not independently quantified, and complete model/interface-state metadata were not preserved for every live response. The study measures associations and descriptive patterns, not causal ranking factors.

Conclusion

Emerging brands can enter AI buyer shortlists before broad mass-market recognition, but shortlist visibility was uncommon in this sample and website Recommendation Readiness alone did not explain which brands were shortlisted. The observed patterns are consistent with a multi-layer, query-dependent evidence environment involving semantic relevance, owned decision-support evidence, external corroboration and query-specific competitive retrieval.

Being searchable is one thing. Being recommendable is another. Being repeatedly shortlisted is harder still.

Suggested citation

Sarkar, S. (2026). Can Emerging D2C Brands Become AI-Recommendable Before They Become Famous? An Exploratory Study of ChatGPT Buyer-Intent Shortlists in India. Digital Platform 271.