Home / Resources / AI visibility by industry
GEO · Original dataAI visibility by industry: our prompt-panel results across 10 Indian sectors
Between July and August 2026 we ran buyer-intent prompt panels for 210 well-known Indian brands across 10 sectors. This page rolls the results into one table you can cite, with the method and limits stated plainly.
Key takeaway: Across 210 well-known Indian brands in 10 sectors, 67% (141) never appeared in AI answers to buyer-intent category questions, and only 10% (21) were both visible and accurately described. Of the 69 brands that did appear, 70% (48) were described with at least one material error. Manufacturing was the weakest sector (72% invisible, 8% pass rate).
The data: 10 sectors, 210 brands
| Sector | Brands tested | Never appeared | % invisible | Appeared | Appeared with errors | Visible and accurate | Pass rate |
|---|---|---|---|---|---|---|---|
| Manufacturing (B2B) | 25 | 18 | 72% | 7 | 5 | 2 | 8% |
| Logistics | 20 | 14 | 70% | 6 | 4 | 2 | 10% |
| Quick-service restaurants | 20 | 14 | 70% | 6 | 4 | 2 | 10% |
| Real estate developers | 20 | 14 | 70% | 6 | 4 | 2 | 10% |
| Banking, finance and insurance | 25 | 17 | 68% | 8 | 5 | 3 | 12% |
| Law firms | 15 | 10 | 67% | 5 | 3 | 2 | 13% |
| Hospital chains | 20 | 13 | 65% | 7 | 5 | 2 | 10% |
| Hotels | 20 | 13 | 65% | 7 | 5 | 2 | 10% |
| Direct-to-consumer brands | 25 | 16 | 64% | 9 | 7 | 2 | 8% |
| Edtech | 20 | 12 | 60% | 8 | 6 | 2 | 10% |
| All 10 sectors | 210 | 141 | 67% | 69 | 48 | 21 | 10% |
Sorted by share of brands that never appeared. "Visible and accurate" equals brands that appeared minus those described with a material error. The percentage that had errors among brands that appeared is 70% overall (48 of 69).
What the numbers say
- Most well-known brands are missing from AI shortlists. Two in three brands that dominate their categories offline never appeared in a category answer in our panels.
- Appearing is not enough. Seven in ten of the brands that did appear were described with at least one material error, such as outdated facts, wrong locations or confusion with a similarly named company.
- Pass rates cluster around 10%. Between 8% and 13% of brands in every sector were both visible and accurate.
- Direct-to-consumer brands had the highest error share. Seven of the nine D2C brands that appeared (78%) carried material errors, despite strong social and marketplace presence.
- Industrial B2B was least visible. Manufacturing had the highest share of invisible brands (72%) and the lowest pass rate (8%), even among listed, export-certified companies.
- Edtech was the most visible sector, but still 60% invisible.
How we measured it
- Brands: 15 to 25 well-known brands per sector, chosen by us for national or regional recognition. Each study lists the brands tested.
- Prompts: three types per brand: category or buyer-intent questions ("best X in India for Y"), direct questions about the brand, and comparison questions.
- Engines: ChatGPT (GPT-4o), Gemini Advanced and Perplexity.
- Visible: the brand appeared, unprompted, in at least one category answer.
- Accurate: direct answers described the brand without a material error on products, geography or status, as judged by our team.
- When: July to August 2026.
A separate cross-sector study of 50 Indian mid-market brands, tested across four engines including Claude, found 34 invisible (68%), 16 that appeared, 9 of those with material errors, and 7 that were both visible and accurate (14%). It is a different sample, so we do not add it to the totals above. See the 50-brand study.
Limits of this data
- It is not a random sample. Sample sizes are small, so differences of a few percentage points between sectors are indicative, not conclusive.
- We designed the prompts, and a different prompt set would produce different numbers.
- AI answers change with model updates, user, location and phrasing. This is a snapshot, not a permanent state.
- "Invisible" means absent from category answers in our panel. It does not mean the brand is never mentioned in any AI answer.
- Accuracy was judged by our team against public information.
- We do not break results out by engine in this roll-up. Where a brand is "invisible", it did not appear in any category prompt on any engine we tested.
Why brands go invisible
The studies point to recurring causes: little or no presence on the reference sites and directories engines cite, brochure-style website copy with nothing quotable, content that only renders with JavaScript, and in four cases of the 50-brand study, AI crawlers blocked by default settings. In some sectors the pattern was specific: restaurant brands described through aggregator listings, hotel and hospital chains collapsed to a single flagship, and delivered projects described as current. The fix is systematic. See the complete GEO guide and how to get cited by ChatGPT.
How to cite this data
Reckona AI (2026), "AI visibility by industry: our prompt-panel results across 10 Indian sectors", https://reckonaai.com/resources/ai-visibility-by-industry-india-prompt-panel-data/. You are welcome to quote the figures with that attribution and the limits above.
Test your own brand
You can run a small version of the same test in minutes with our free AI Rank Tracker, which builds a panel of buyer prompts for your brand, and score what ChatGPT, Gemini, Perplexity and Claude say. Our free audit runs the full monthly panel for you.
Frequently asked questions
What percentage of Indian brands are invisible in AI answers?
In Reckona AI's 2026 prompt panels, 67% of 210 well-known Indian brands across 10 sectors (141 brands) never appeared in an AI answer to a buyer-intent category question on ChatGPT, Gemini or Perplexity. The share ranged from 60% in edtech to 72% in manufacturing.
How many brands are described accurately when they do appear?
Of the 69 brands that appeared, 48 (70%) were described with at least one material error. Only 21 of 210 brands (10%) were both visible and accurately described.
Which industry has the lowest AI visibility in India?
In this data, manufacturing: 18 of 25 brands (72%) never appeared and only 2 (8%) were visible and accurate. Direct-to-consumer brands had the highest share of errors among those that appeared (7 of 9).
How was the data collected?
Our team ran three prompt types (category, direct and comparison) for 15 to 25 well-known brands per sector across ChatGPT (GPT-4o), Gemini Advanced and Perplexity in July and August 2026. A brand counted as visible if it appeared unprompted in at least one category answer.
Can I use these statistics?
Yes, with attribution to Reckona AI and the page URL, and with the limits in mind: samples are small, prompts were chosen by us, and AI answers change over time.
Want this done for your brand?
Start with a free audit: a written scope and a baseline of how AI assistants describe your brand today.
Get my free audit →