// ai visibility
AI visibility, without the guesswork
Why answer engines skip your business, what genuinely helps, and what's being sold to you that does nothing. Every claim here is sourced, and where I don't have data, I say so instead of inventing a statistic.
AI visibility has become a market, and most of what's sold in it is confident nonsense: invented percentages, guaranteed placements, and file formats nobody reads. These pages are the opposite of that. Where there's evidence, it's linked. Where I've measured something myself, the raw numbers are published. Where I haven't, I say so.
The two things worth knowing first
- AI crawlers don't run JavaScript. Per Vercel's crawler study, none of the major AI crawlers execute JavaScript, while Googlebot does. That single asymmetry explains most cases of "we rank on Google but ChatGPT has never heard of us".
- Nobody can guarantee an AI recommends you. There's no dial. There's retrieval, training data, and probability. You can be readable and corroborated; you cannot be guaranteed.
- Most of what's sold as a fix does nothing. Google says it does not support llms.txt, and Ahrefs found 97% of valid llms.txt files got zero AI requests in May 2026. Google also deprecated FAQ rich results in 2023; by 2026 they had stopped appearing. Spend the effort on being readable instead.
What I've actually measured
The AI Visibility Index is a published scan of real Indian business websites, named, with every score and the method attached. Two verticals are live so far, 64 sites between them. That's the kind of claim this site makes: narrow, checkable, and boring enough to be true.
120
questions customers ask AI
Every question mapped to what the AI needs from your site and the exact fix, across 8 categories. Filterable, free, no signup.
open the question index →
35
Indian clinic sites scanned
Median 87/100, every published score named and re-runnable. The 4 sites with no structured data score a median of 54.
open the clinic index →
The interactive map: 632 sites, 28 industriesEvery scan on one page as a colour-coded map. Median 87 out of 100 and 81% missing FAQ markup across the whole corpus. Hover any dot; recolour by grade, structured data, or FAQ. The fastest way to see the pattern.
Case study: a clinic went 81 to 92 in 3 daysOne real dental clinic, before and after. Two structural fixes (a page title and a meta description) moved its AI readability score from 81 to 92, B to A, in three days. Every number from a re-runnable scan.
The ranking: 28 verticals, 632 sitesEvery scan in one ranked table, from dietitians and immigration consultants at a median 91.5 down to gyms at 53.5. Structured data is 73% adopted and mostly solved; FAQ markup, at 14%, is the blind spot every vertical shares.
Law firm scan: 29 sites, median 77/10093% have no LegalService or Attorney schema type. Firms with structured data score a median of 89; firms without, 55. A 34-point gap.
Clinic study: what 16 sites add up toThe 16 published clinic rows, aggregated: median 77, not one A, and structured data as the dividing line (median 54 without it, 80.5 with). Every number recomputed from the index.
Hotel study: 25 Indian hotels scannedMedian 76, a 27-point structured-data gap, and all 25 missing FAQ markup. Why AI travel answers cite the OTA, not the hotel.
Restaurant study: 15 Indian restaurants scannedMedian 79, a 31.5-point structured-data gap (the widest of any vertical), and all 15 missing FAQ markup. Why AI sends diners to Zomato, not your site.
IVF and fertility study: 14 clinics scannedMedian 90, the highest of any vertical. All 14 publish structured data, the first fully solved vertical, and FAQ markup leads the field here at 6 of 14, though still under half.
Hair transplant study: 13 clinics scannedMedian 89, near the top of every vertical. Structured data is near universal here and 4 sites even have FAQ markup, a first. What a competitive paid-search category already gets right.
Dermatology study: 17 clinics scannedMedian 89. 13 of 17 publish structured data and the sites without it sit far lower, but only 1 of 17 has FAQ markup, the widest blind spot of any clinic vertical.
Cosmetic surgery study: 15 clinics scannedMedian 87. Schema is near universal, so the gap moves to on-page basics: only 7 of 15 have a clean title tag and just 1 of 15 has FAQ markup.
Preschools and daycare study: 14 sites scannedMedian 87, a solid mid-table vertical (9 of 14 grade A). Structured data is near universal (13 of 14), but only 3 of 14 carry the FAQ schema that answers what parents ask AI: fees, age, timings, safety, ratios, meals.
Eye and LASIK study: 18 clinics scannedMedian 87, among the strongest verticals. Structured data is near universal (17 of 18), but only 1 of 18 has the FAQ markup that answers LASIK and cataract questions.
Gyms and fitness studios study: 16 sites scannedMedian 53.5, the lowest-scoring vertical measured so far: 5 of 16 scored an F. Structured data splits the field 87 vs 39, and boutique studios are near-invisible to AI.
Ayurveda and panchakarma study: 18 sites scannedMedian 87, among the strongest verticals measured: 10 of 18 scored an A. Structured data is nearly universal (16 of 18), so the gaps that remain are FAQ markup (2 of 18) and a single clear H1 (4 of 18).
Law firms study: 18 sites scannedMedian 59.5, one of the weakest verticals, split hard by firm size (national 83, mid-size and boutique far below). Structured data (7 of 18) is the dividing line, and FAQ markup sits at zero of 18.
Study-abroad consultants study: 20 sites scannedMedian 90, one of the strongest verticals measured (level with IVF) and the first non-medical category to reach that level: 14 of 20 scored an A, two a perfect 100. Structured data nearly universal (18 of 20); FAQ markup the gap left (5 of 20).
CA and accounting firms study: 16 sites scannedMedian 87, a top professional-services vertical (against law firms at 59.5), lifted by the digital-first compliance platforms. Structured data (11 of 16) is the divider; FAQ markup, the cheapest fix, sits at 1 of 16.
Digital agencies study: 173 sites scannedThe largest scan yet, and an on-brand one: median 89 places the agencies who sell findability 5th of 28 verticals. But it is not uniform. One in five scored below a B (range down to 23), and only 31 of 173 carry FAQ schema, the vertical's clearest gap.
Dietitian and nutrition clinics study: 16 sites scannedMedian 91.5, the strongest vertical scanned to date (tied top of the 28-vertical ranking). Structured data is near-universal (15 of 16) and 14 of 16 grade A. The lone gap is FAQ schema, on just 3 of 16, despite dietitians answering the most FAQ-shaped questions of any profession.
Pediatric clinics study: 17 sites scannedMedian 77 but a split field: the 9 clinics with structured data sit at a median 92, the 8 without collapse to about 54, a 38-point gap on one signal. Despite parents asking the most FAQ-shaped questions in medicine (fever, vaccine timing, dosing), only 3 of 17 carry FAQ schema.
Interior designers study: 15 sites scannedMedian 78, bimodal: the tech-enabled platforms hit a median 92 (one perfect 100) while design-led studios trail. Structured data splits the field 87 vs 23, a 64-point gap, the widest measured. Visual-craft firms build the least machine-readable sites.
Wealth management firms study: 20 sites scannedMedian 88, one of the strongest verticals measured: 14 of 20 scored an A and none failed. Structured data still sorts the bottom (the 3 without it scored 57, 57 and 61); firm size does not predict readability (a solo fee-only planner hit 100, a large family office 57). FAQ markup is the gap left, on 4 of 20.
Immigration consultants study: 18 sites scannedMedian 91.5, the strongest vertical measured to date: 14 of 18 grade an A, two perfect 100s. A business built on answering eligibility, points and timeline questions builds pages an AI reads well. Structured data still sorts the bottom three (43, 53, 58); FAQ markup, the highest adoption seen, is on 7 of 18.
Homeopathy clinics study: 18 sites scannedMedian 86 but a two-tier field (mean 77.8): the 11 clinics with structured data sit at a median 92, the 7 without it collapse to 59, close to 40 percent of the vertical. In one of the most question-driven fields in medicine, only 2 of 18 carry FAQ schema.
Coworking spaces study: 15 sites scannedMedian 87, a uniformly competent vertical (no failing grades). The first vertical where structured data is basically solved (14 of 15), so it flips the usual story: the gap is one level up. Only 2 of 15 carry FAQ schema and 9 of 15 a single clean H1.
Physiotherapy and rehab clinics study: 15 sites scannedMedian 87, one of the stronger verticals. Structured data is near universal (12 of 15) and standalone clinics score almost as well as the chains. The gap that remains is FAQ markup, on 3 of 15.
Veterinary and pet hospitals study: 15 sites scannedMedian 66, the most polarized field yet: 7 sites score 87 or higher, 8 score 66 or below, zero B grades. Structured data splits them 90.5 vs 46, and standalone hospitals score lowest.
Diagnostic and pathology labs study: 18 sites scannedMedian 79, range 23 to 96. Structured data splits the field 83 vs 46, only 5 of 18 have a clean H1, and the regional labs a "near me" search should surface score lowest.
Salons and spas study: 14 sites scannedMedian 80.5, 20th of 28 and one of the weaker verticals: 4 of 14 scored below a B (down to 44). Structured data is not universal (10 of 14) and not one site (0 of 14) has FAQ schema, despite salons fielding price, timing and appointment questions constantly.
Coaching and edtech study: 21 sites scannedMedian 83, the widest spread yet (23 to 92). The biggest JEE/NEET brands score lowest, UPSC institutes highest, and 0 of 21 have FAQ markup.
Cosmetic dentistry study: 14 clinics scannedMedian 87. 13 of 14 publish structured data, but only 2 of 14 have FAQ markup, the same fix the real 81 to 92 dental case study used to reach an A.
US, Canada and Australia: 24 sites scannedHalf scored an A, but the sites without structured data sat 30 points lower. The gap is legibility, not market maturity.
Startup study: 36 funded startups scannedMedian 81, a third below a B, 88% with no FAQ schema and 42% with no structured data at all. Being funded does not make you readable to AI, and 12 of the 36 blocked the automated read outright.
Hotel and restaurant indexes will be published here as the scans land.
The questions customers ask AI, by vertical
Each of these takes the real questions people type into an answer engine in one industry and maps every one to the specific on-site fix that lets the answer be you.
15 questions dental patients ask AIThe 15 questions patients ask ChatGPT, and the exact on-site fix that lets the answer be your clinic. Grounded in the clinic scan (median 54 without structured data, 80.5 with).
15 questions hotel guests ask AIThe prompts travelers type to pick a hotel, each with its fix. Grounded in a scan of 25 hotel sites: median 76, a 27-point structured-data gap.
15 questions restaurant guests ask AIThe prompts diners type to pick a restaurant, each with its fix. Grounded in a scan of 15 restaurant sites: median 79, a 31.5-point structured-data gap, all 15 missing FAQ markup.
15 questions fertility patients ask AIThe heavy questions IVF patients ask (success rates, cost, IVF vs IUI), each with its fix. Built on named-author credibility and honest sourced data, not marketing spin.
15 questions hair transplant patients ask AIThe questions patients type (cost, grafts, timeline, pain), each with its on-site fix. Grounded in a scan of 13 clinic sites: median 89, near the top of every vertical, yet only 4 of 13 have FAQ markup.
15 questions dermatology patients ask AIThe questions derm patients type (acne cost, pigmentation, PRP, laser), each with its on-site fix. Grounded in a scan of 17 clinic sites: median 89, but only 1 of 17 has the FAQ markup that answers them.
15 questions cosmetic surgery patients ask AIThe questions patients type (rhinoplasty cost, recovery, scars, financing), each with its on-site fix. Grounded in a scan of 15 clinic sites: schema near universal at 14 of 15, but FAQ markup on just 1 of 15.
15 questions eye and LASIK patients ask AIThe questions patients type (LASIK cost, safety, recovery, cataract lens types, financing), each with its on-site fix. Grounded in the 18-clinic eye scan: schema near universal, but FAQ markup on just 1 of 18.
15 questions coaching and edtech students ask AIThe questions students and parents type (fees, batch dates, results, scholarships, refund policy), each with its on-site fix. Grounded in the 21-site coaching scan: 0 of 21 have FAQ markup and only 6 have a clean H1.
15 questions patients ask AI before choosing a diagnostic labThe questions patients type (test price, home collection, NABL accreditation, report turnaround, timings), each with its on-site fix. Grounded in the 18-lab scan: only 3 of 18 have FAQ markup and only 5 have a clean H1.
15 questions pet owners ask AI before choosing a vetThe questions owners type (emergency hours, vaccination cost, home visits, surgery, is the clinic open now), each with its on-site fix. Grounded in the 15-hospital scan: only 1 of 15 has FAQ markup and only 7 have a clean H1.
15 questions people in pain ask AI before booking a physioThe questions patients type (session cost, home visits, back pain, sciatica, sports injury, post-surgery rehab), each with its on-site fix. Grounded in the 15-clinic scan: only 3 of 15 have FAQ markup, the clearest remaining gap.
15 questions people ask AI before joining a gymThe questions people type (membership cost, 24 hour access, personal trainer, trial pass, timings, classes), each with its on-site fix. Grounded in the 16-gym scan, the lowest-scoring vertical: only 1 of 16 has FAQ markup.
15 questions people ask AI before booking an ayurvedic treatmentThe questions patients type (which panchakarma, BAMS or MD doctors, authentic Kerala lineage, program length, cost, residential or outpatient), each with its on-site fix. Grounded in the 18-site scan: only 2 of 18 have FAQ markup.
15 questions people ask AI before hiring a law firmThe questions clients type (practice areas, M&A in Mumbai, SEBI matters, fees, jurisdiction, free consultation), each with its on-site fix. Grounded in the 18-firm scan: zero of 18 have FAQ markup.
15 questions people ask AI before choosing a study-abroad consultantThe questions students and parents type (best country for my profile, service fees, visa success rate, scholarships, IELTS prep, refund policy), each with its on-site fix. Grounded in the 20-site scan: only 5 of 20 have FAQ markup.
15 questions people ask AI before hiring a CA or accountantThe questions founders type (best CA for a startup, ITR and audit fees, GST and company registration, virtual CFO, tax notices), each with its on-site fix. Grounded in the 16-firm scan: only 1 of 16 has FAQ markup.
15 questions people ask AI before hiring an interior designerThe questions homeowners type (2BHK cost, price per square foot, timeline, modular kitchens, warranty, EMI, turnkey scope), each with its on-site fix. Grounded in the 15-firm scan: only 3 of 15 have FAQ markup.
15 questions people ask AI before hiring a financial advisorThe questions investors type (SEBI registration, fee-only vs commission, minimum portfolio, fiduciary duty, custody, NRI service), each with its on-site fix. Grounded in the 20-firm scan: only 4 of 20 have FAQ markup.
15 questions people ask AI before hiring an immigration consultantThe questions applicants type (Canada or Australia PR eligibility, CRS cutoff, fees, timeline, ICCRC or MARA registration, refund on refusal), each with its on-site fix. Grounded in the 18-firm scan: only 7 of 18 have FAQ markup.
15 questions people ask AI before choosing a homeopathy clinicThe questions patients type (does it work for thyroid or hair loss, safe for kids, combine with allopathic medicine, cost, timeline, qualification), each with its on-site fix. Grounded in the 18-clinic scan: only 2 of 18 have FAQ markup.
15 questions people ask AI before choosing a dietitianThe questions people type (consult fee, online option, PCOS/thyroid/diabetes plans, vegetarian Indian food, RD credentials, follow-ups, refunds), each with its on-site fix. Grounded in the 16-clinic scan: dietitians score highest of any vertical, yet only 3 of 16 have FAQ markup.
15 questions parents ask AI before choosing a pediatric clinicThe questions parents type (clinic timings, consultation fee, vaccine schedule and stock, Sunday and emergency hours, teleconsult, the doctor's qualifications), each with its on-site fix. Grounded in the 17-clinic scan: only 3 of 17 have FAQ markup.
15 questions people ask AI before choosing a coworking spaceThe questions founders type (hot-desk price, day pass, locations, meeting-room rates, GST or virtual office, lock-in), each with its on-site fix. Grounded in the 15-brand scan: structured data is near-universal, but only 2 of 15 have FAQ markup.
Start here
Why ChatGPT recommends your competitor instead of youChatGPT names your competitor because of where it reads about them, not because their site is better. The real mechanism, how to check, and what you can change.
Why your business doesn't show up on ChatGPTFour real reasons AI answer engines never mention your business, one popular myth that does nothing, and the specific fix for each. No invented statistics.
AI visibility for hotelsAI travel answers cite OTAs, not hotel websites. Why that happens, what Hotel schema does, and the booking-widget problem that hides your rooms from crawlers.
AI visibility for dentistsI scanned 35 Indian clinic sites. Most were fine: median 87/100. The four with no structured data scored a median of 54. What that means for dental practices.
AI visibility for law firmsIndian advocates can't advertise. That makes AI visibility a different problem: factual machine-readable presence, not promotion. What BCI Rule 36 permits.
AI visibility for restaurantsMost restaurant menus are images, PDFs or JavaScript widgets that AI crawlers cannot read. Why Zomato gets cited instead of you, and how to fix the menu problem.
Report Card: free, no signup
Paste your URL. A 0-100 answer-engine readability score and a ranked list of what to fix, in about ten seconds. Nothing is stored. Prefer to see the fixes first? Four before/after sample rebuilds are here.
Grade my site free →
Frequently asked questions
What is AI visibility?
Whether AI answer engines like ChatGPT, Perplexity and Google's AI surfaces can read your website, understand what your business is, and use it when answering a relevant question. It is separate from ranking on Google, because the crawlers behave differently. Most notably, AI crawlers do not execute JavaScript while Googlebot does.
Can anyone guarantee ChatGPT will recommend my business?
No. There is no ranking dial and no submission form. What can be done honestly is to make your business readable and corroborated, which raises the odds. Anyone guaranteeing placement is selling certainty they do not have.
Is llms.txt worth adding?
Not for visibility. Google has said it does not support llms.txt and has no plans to, and Ahrefs found 97% of valid llms.txt files received zero AI requests in May 2026. It is harmless but it is not the reason you are invisible.
How do I check my own site?
The Report Card is free, needs no signup, and scores any URL out of 100 with a ranked fix list in about ten seconds. Nothing is stored.