What 334 AI Citations Reveal About Product Discovery in India

While reviewing recent AI Search experiments from Indian practitioners, India SEO Community came across a study by Growgence examining where ChatGPT and Perplexity source information when answering commercial recommendation questions.

The original finding was striking. Independent review media accounted for around 27% of citations in US-focused answers, compared with only 2% in India-focused answers. SEO and affiliate-style blogs, meanwhile, appeared at similar rates in both markets.

That raised a question worth exploring: if review publications were largely absent from the India-focused answers, which sources were AI systems using instead?

We reached out to Growgence for the methodology and underlying breakdown. The team shared the results by market, category and AI engine, along with the limitations of the experiment. The additional data made the original finding more nuanced and more useful for SEOs.

How the experiment worked

Growgence selected six commercial categories: skincare and personal care, eyewear, exam preparation, payments and payroll, supplements and protein, and wireless audio.

The team used buyer-style questions such as “What are the best…” and “Which are the best…”, keeping the India and US versions structurally similar. Exam preparation was the exception because the underlying exams needed localisation, with JEE and NEET used for India and SAT and AP for the US.

The experiment therefore was not measuring which brands ranked highest. It was examining the information sources sitting behind AI-generated buying answers.

The biggest difference was not SEO blogs

The first useful finding appears when the entire citation mix is compared. Independent review media represented 27.2% of US citations and only 2.3% of India citations. The difference is substantial, but another number helps explain why the finding matters. SEO and affiliate blogs accounted for 44.8% of India citations and 41.4% of US citations. Their presence was about the same in both markets.

The source mix diverged much more strongly around brand and retail properties. Brand-owned sites accounted for 26.2% of India citations versus 18.5% in the US, while marketplaces accounted for 14.5% versus 3.7%. Together, brand-owned sites and marketplaces represented 40.7% of India citations, compared with 22.2% in the US sample.

The experiment does not establish that brand and marketplace sources directly replaced review publications. It does show that the information mix supporting India-focused recommendations was materially different. For SEOs, this is more actionable than the review-media gap alone.

The review-media gap was concentrated in two categories

The review-media gap highlighted earlier also turned out to hide an important detail. Of the 44 US review-media citations, 22 came from wireless audio and another 14 from eyewear. Those two categories alone accounted for 81.8% of all US review-media citations in the study. Wireless audio provides the clearest example: review publications contributed 22 of 29 US citations, or 75.9%, while for India they contributed only one of 30, or 3.3%. The India answer set instead leaned on marketplaces and SEO blogs, with 13 citations each. Eyewear showed a related pattern: review media represented 48.3% of US eyewear citations, while brand-owned websites represented 60% of the India eyewear citations.

This suggests the aggregate India-versus-US comparison should not be interpreted as a universal market rule. The difference was particularly strong in categories where the US has an established specialist review ecosystem. For SEO research, that distinction matters. A broad GEO benchmark might describe what happens across several industries while hiding how sourcing actually works in the one industry a team cares about.

There was no single India sourcing pattern

Looking only at the India data reinforces this point. SEO/blog content dominated payments, exam preparation and skincare. Brand-owned sources dominated eyewear and supplements. Wireless audio was split between marketplaces and SEO/blog content.

The practical implication is that “What sources does AI use in India?” is too broad a question for an SEO strategy. A better research question is: which sources do AI systems use when answering the commercial questions in our category? An eyewear company following the overall India average might focus heavily on external blogs because those sites accounted for 44.8% of the full India dataset. Yet within eyewear, 60% of citations came from brand-owned sites. Category-level evidence changes the strategy.

ChatGPT and Perplexity produced different source mixes

Growgence found another difference at the engine level. Perplexity produced 120 India citations, while ChatGPT produced 52. Growgence noted that Perplexity generally returned close to 20 sources per answer, while ChatGPT returned fewer. So raw citation counts should not be compared across engines as if both systems expose sources the same way. The composition within each engine is still informative.

Marketplace, LinkedIn, earned media and reference citations that Growgence didn’t break out individually per engine are grouped under “Other sources.”

Among ChatGPT’s 52 India citations, brand-owned sites, marketplaces and SEO/blog sources each appeared 16 times. There were no Reddit or independent review-media citations at all. Perplexity’s 120 India citations included 61 SEO/blog citations, 29 brand-owned citations, 10 Reddit citations and four review-media citations. Every Reddit citation and every review-media citation on the India side therefore came from Perplexity.

For SEO teams, combining ChatGPT, Perplexity and other AI systems into a single “AI visibility score” risks hiding differences in how each system retrieves and presents sources.

What SEOs should take from the experiment

The most useful learning is methodological. AI Search research should move beyond checking whether a brand appears in a response. The sources behind the response often provide more information about where visibility is coming from and where an SEO team should investigate next.

1. Trace competitor recommendations back to their sources

   If a competitor is repeatedly recommended, the next question should be where the supporting information originates. A recommendation supported by the competitor’s own product pages presents a different problem from one supported by Amazon, an independent review publication, an affiliate comparison or a Reddit discussion.

2. Run the analysis at category level, not account level

   Growgence’s data also makes a strong case for category-level analysis. The six Indian categories produced materially different source patterns, so generic GEO recommendations risk sending teams toward source types that have little influence within their own commercial queries.

3. Treat brand-owned pages as retrieval surfaces, not just conversion pages

   Brand sites represented 26.2% of the full India citation pool and more than half of the citations in eyewear and supplements. For SEOs, this makes clear product descriptions, specifications, comparisons, use cases, entity information and factual consistency, not just an on-site conversion concern. These pages are also part of the information environment that retrieval systems read.

4. Audit marketplace listings the same way

   Marketplace listings deserve similar scrutiny for ecommerce brands. They represented 14.5% of India citations overall and 43.3% within wireless audio. Product attributes, descriptions, specifications, categories and consistency between retailer listings and the main website belong in an AI Search audit.

5. Don’t treat Reddit as a shortcut

   Reddit represented 5.8% of India citations and 5.6% of US citations, almost identical shares. On the India side, its presence came entirely through Perplexity. The evidence here does not support treating Reddit as a universal shortcut to AI visibility.

6. Treat this as a snapshot, not a benchmark

   Growgence tested each market-category-engine combination once, so the study captures a snapshot of sourcing behaviour, not its stability. A stronger SEO measurement programme would repeat important commercial prompts, record every citation, classify the sources, and track which domains and source types persist across runs and over time.

The conclusion

Growgence’s original experiment started with a simple comparison between India and the US. The deeper breakdown leads to a more useful conclusion: in this 25 August 2026 snapshot, India-focused buying answers contained far fewer independent review-media citations and substantially more brand-owned and marketplace citations than the US sample. Yet the difference was not consistent across categories, and ChatGPT and Perplexity did not expose the same source patterns.

For SEOs, the lesson is therefore not to chase one source type. It is to map the information ecosystem around the commercial questions that matter to the business: identify which source types appear, which domains keep coming up, where competitors get their support, how the pattern shifts by AI engine, and whether those observations hold up under repeated tests. Only then does the discussion move from generic GEO advice to evidence-led AI Search strategy.