Skip to main content
Cited publishes aggregate benchmark ranges for AI visibility metrics based on observed data from Indian brands across 8 categories. This page documents the sample, methodology, and caveats behind those benchmarks.

​
Data sources

Benchmarks combine data from two sources:
  • Cited Index — a public leaderboard of 253 Indian brands across 8 categories (CRM & Sales, Digital Payments, Credit Cards, Online Learning, HR & Payroll, Audio & Wearables, Skincare & Beauty, Travel & Luggage). Each brand is tracked against 25 category-specific prompts across 4 AI platforms (ChatGPT, Perplexity, Gemini, Claude). The Index is refreshed monthly.
  • Pilot brands — 4 brands onboarded for platform validation, tracked daily across up to 5 AI platforms (including Grok). Pilot brand data provides higher-frequency signal but a much smaller sample.
All published numbers are aggregate percentiles. No individual brand metrics appear anywhere in the benchmark data.

​
What we measure

Five metrics are tracked. Three have sufficient data for benchmark publication: Two have limited data:
  • Sentiment — categorical only from the Cited Index (positive, mixed, negative). Numeric sentiment scores are available for pilot brands but the sample is too small for percentile distributions.
  • Citation rate — Perplexity-only with 4 pilot brands. Benchmark-grade citation rate data is planned for Q3 2026 when the Cited Index adds citation tracking.

​
How the percentiles are computed

For each metric, we compute the percentile distribution across all brands in the dataset. The reported percentiles are p10, p25, p50 (median), p75, and p90. For overall benchmarks, all 281 brand observations are included. For category-specific benchmarks, percentiles are computed within each category using only the brands in that category. Share of voice is computed as each brand’s visibility score divided by the sum of all visibility scores in its category, expressed as a percentage. This means SoV percentiles reflect how concentrated or distributed mentions are within a category — a high p75 indicates a long-tail distribution with a few dominant brands.

​
Caveats

  • Benchmarks reflect the Indian market as of March 2026. Other markets and B2B categories may show different distributions.
  • Categories with fewer than 25 brands have noisier percentile estimates. Skincare & Beauty (21 brands), Travel & Luggage (21), and Audio & Wearables (22) are at the lower end.
  • Grok was added to the Cited platform in April 2026 and is not yet included in the Cited Index. Grok benchmarks will appear in the next monthly refresh.
  • The Cited Index is refreshed monthly; pilot brand data is daily. Cross-source comparisons should account for this frequency difference.
  • Sentiment benchmarks are categorical only from the Cited Index. The majority of brands (86.5%) are classified as “mixed” sentiment, which limits the discriminating power of this metric at the aggregate level.
  • Citation rate is Perplexity-only and sample-limited. Excluded from published benchmarks until Q3 2026.
  • Refreshed monthly alongside the Cited Index.

​
What is NOT in the benchmarks

  • Individual brand numbers. All data is aggregate — no brand name is tied to a metric value.
  • B2B SaaS categories. Only 2 pilot brands are B2B SaaS, which is insufficient for category benchmarks.
  • Non-Indian markets. The Cited Index currently covers Indian brands only. International benchmarks are planned.
  • Pre-March 2026 data. The platform launched recently; historical baselines do not exist.

​
Last refreshed

These benchmarks were generated on April 12, 2026 using the March 2026 Cited Index edition and pilot brand data through April 11, 2026.