Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.getcited.in/llms.txt

Use this file to discover all available pages before exploring further.

Cited publishes empirical benchmark ranges for AI visibility metrics — mention rate, share of voice, and average position — computed from the Cited Index, a monthly snapshot of Indian brands across 8 categories. This page documents the data source, methodology, and caveats behind those benchmarks.

​
The data source

Benchmarks are computed from the Cited Index — a monthly measurement of how 253 Indian brands appear in AI-generated answers across ChatGPT, Perplexity, Gemini, and Claude. Each brand is measured using a fixed set of 185 non-branded category queries distributed across 8 categories. The current edition reflects data collected during the March 2026 pipeline run. The next edition is planned for early May 2026.

​
What “non-branded category query” means

A non-branded category query is a question a customer might ask an AI without naming any specific brand. Examples:
  • “best travel bags under ₹5000 in India”
  • “top HR software for Indian startups”
  • “which skincare brands are good for oily skin”
When the AI answers, some brands get mentioned, others don’t. The Cited Index measures which brands appear in which categories — without ever telling the AI what to say.

​
Why this matters

Non-branded category queries are the hardest test of AI visibility. If a customer has to name your brand for the AI to mention you, you have no organic presence inside AI search. You only appear when someone already knows you exist. Most brands fail this test. The median Indian brand in the Cited Index has a mention rate of just 8% — meaning the AI mentions them in roughly 1 out of every 12 relevant category queries. Even top-quartile brands only hit 16%. This is the real state of AI visibility for Indian D2C and B2B SaaS in early 2026, and it’s the opportunity GEO exists to address.

​
What these benchmarks do NOT measure

It’s important to be explicit about what these benchmarks exclude:
  • Branded queries — “is Brand X a good travel bag” or “Brand X reviews”
  • Comparison queries — “Brand X vs Brand Y” or “Brand X alternatives”
  • Custom prompt libraries — brand-specific prompt sets generated by Cited’s query intelligence for individual customers
All three of the above typically produce much higher mention rates than non-branded category queries, because the AI is being prompted to talk about a specific brand rather than being asked to surface one organically.

​
Why your dashboard may show higher numbers

If you run Cited on your own brand, your dashboard will often show mention rates significantly higher than these benchmarks — sometimes 2x to 5x higher. This is expected. Your dashboard uses a custom prompt library generated specifically for your brand. That library is a richer mix that includes category queries, comparison queries, use-case queries, and semi-branded queries — designed to give you the fullest picture of how AI describes your brand in real customer conversations. The Cited Index benchmark, by contrast, uses only non-branded category queries — a narrower, harder measurement. Both numbers are correct. They measure different things. When comparing your dashboard to these benchmarks, the useful question is: “How do I perform on the hardest test?” If your non-branded category mention rate is above the p75 benchmark for your category, you are genuinely visible in AI search without any tailwind from branded queries.

​
How percentiles are computed

For each metric, Cited computes percentile distributions (p10, p25, p50/median, p75, p90) across all 281 brand-category observations in the Index. Brands appearing in multiple categories are counted as independent observations for overall statistics and deduplicated within individual category statistics. Per-category percentiles are only published for categories with at least 20 brands. Categories with smaller samples are flagged as having noisier estimates.

​
Caveats

  • Benchmarks reflect the Indian market as of March 2026. Other markets may show different distributions.
  • Categories with fewer than 25 brands have noisier percentile estimates: Skincare & Beauty (21), Travel & Luggage (21), Audio & Wearables (22).
  • The Cited Index currently tracks 4 platforms: ChatGPT, Perplexity, Gemini, Claude. Grok is tracked on the main Cited platform but not yet integrated into the Cited Index pipeline — will be added in a future edition.
  • Sentiment and citation rate benchmarks are not included in this edition. Sentiment classification produces too much “mixed” output to yield meaningful percentiles. Citation rate data is too sparse. Both will be added when data quality improves.

​
Refresh cadence

The Cited Index is refreshed on a monthly cadence. Each edition is dated in the methodology page and in the downloadable benchmark data file.
  • Mention rate — the single most important visibility metric
  • Share of voice — brand’s share of total mentions within a category
  • Average position — where brands typically appear in citation lists
  • The full Cited Index on getcited.in — leaderboard format, browsable by category