The State of Brands in AI · India Edition
The world's second-largest AI market,watched for 63 straight days.
50 unaided buyer-language prompts. ChatGPT, Gemini, Google AI Mode. Every night from May 19 to July 21, 2026, across 10 of India's most-funded consumer categories. 90,132 answers on share of voice, source pool, sentiment, and day-over-day movement. The study has concluded; every edition stays online, and the complete findings live in the final report. Open methodology. Public data. Cite us when you use it.
10
Verticals tracked
96+
Brands across categories
90,132
AI answers collected
63 days
May 19 to Jul 21, 2026
Ola Electric
Bajaj Chetak
TVS iQube
Hero Vida
Simple One
River
Ampere
Revolt
Mahindra BE 6
Zepto
Flipkart Minutes
Minimalist
The Derma Co
Zerodha
Groww
Yoga Bar
The Whole Truth
PhysicsWallah
Adda247
Razorpay
Cashfree
MakeMyTrip
ixigo
Acko Health
PlumOla ElectricBajaj ChetakTVS iQubeHero VidaSimple OneRiverAmpereRevoltMahindra BE 6ZeptoFlipkart MinutesMinimalistThe Derma CoZerodhaGrowwYoga BarThe Whole TruthPhysicsWallahAdda247RazorpayCashfreeMakeMyTripixigoAcko HealthPlumThe final report · every finding from 63 days
In 4 of 10 categories, the AI leader never changed once in 63 days.
The complete record of the study is now published: who each engine recommends, who it cites, why Reddit out-cited every news publisher in India, and why AI visibility behaves like a moat rather than a lottery. Every number is final and verified against the full 90,132-answer corpus.
Biggest movers on the final day · 21 July 2026
The 10 verticals
India's most-funded consumer categories, watched on AI for 63 days.
50 prompts × 3 engines × 63 days
Methodology
The prompts. The engines. The math. All public.
Unaided buyer prompts
50 prompts per vertical. Zero brand names inside the prompt itself. The kind of question a real buyer in India would type. Branded prompts corrupt share-of-voice math.
Three engines, every night
ChatGPT (gpt-5-5), Gemini 3 Flash, Google AI Mode. Same prompts. India location. Run nightly at 21:00 IST via DataForSEO's LLM scraper endpoints for the full 63-day study.
Three primary metrics
Share of voice (% of prompts naming the brand). Source pool (cited domains, per engine). Sentiment (positive / neutral / cautious / negative).
Daily delta surfacing
Every brand's day-over-day movement was computed across all 621 editions. Movements over 2pp triggered an editorial review for that day's 'what changed' narrative.
Questions