Query fan-out ratio — ChatGPT Ads India glossary
In plain English: if one question about a fertility clinic actually breaks into four or five separate things ChatGPT could answer — and separately auction — how many auction opportunities does a specialty really generate per month? The query fan-out ratio is the number that answers that, and it's the input a raw keyword-volume estimate can't give you.
Definition
The query fan-out ratio is a category-level or specialty-level average expressing how many distinct sub-intent branches a typical opening query splits into once ChatGPT's intent-classification layer processes it, before ad eligibility or auction participation is even assessed. Technically, it is calculated by sampling representative queries for a category, running them (or simulating them) through the fan-out step, counting the resulting branches, and averaging across the sample — a ratio of 1.0 would mean queries in that category rarely fan out at all, while a ratio of 4-5, common in Indian healthcare specialties, means each opening query is actually generating four to five separate auctionable moments across a conversation. This ratio is the multiplier that converts raw query-volume estimates into a realistic count of biddable inventory moments, and it varies significantly by specialty, city tier and even by whether the query is phrased with a location.
Why it matters for Indian healthcare marketers
Healthcare specialties in India tend to carry some of the highest fan-out ratios ICG has observed across any advertiser category on ChatGPT, and fertility, cardiac and aesthetic-medicine queries in particular routinely fan out into four or five sub-intent branches from a single opening question — clinical-informational, cost, provider-comparison, location-specific and booking-transactional branches can all spin off the same initial query within two or three turns. A marketer who budgets against raw query-volume estimates — treating "IVF clinic Mumbai" as one biddable unit the way it would be treated as one keyword in Search — will size their budget against a fraction of the actual auction activity their specialty generates, and will be outbid or simply absent from branches they never accounted for.
But a high fan-out ratio is not the same as a high eligible-inventory count, and that gap is where healthcare marketers most often misjudge the number. NMC Section 6 and ASCI Chapter III compliance filters routinely exclude the clinical-informational branch from paid eligibility entirely for regulated specialties, which means a fan-out ratio of 4.5 for a fertility account might translate to only 2-2.5 branches that are actually legally biddable — the informational branches remain reachable only through organic citation strategy, not sponsored response. Marketers who quote or budget against the raw fan-out ratio without subtracting the compliance-ineligible share will overestimate the addressable inventory their spend can actually reach, and will be disappointed by delivery volume relative to what the raw ratio implied.
Because this metric is specific to the ChatGPT Ads India format and barely 90-180 days into wide market adoption, almost no agency serving Indian healthcare brands has built specialty-level fan-out ratios into their planning process yet — most are still quoting budgets off Search-style keyword-volume logic, which is a fundamentally different unit of measurement for a conversational ad format.
How ICG uses/measures/handles it in a live engagement
ICG calculates a specialty-and-city-specific fan-out ratio for every new ChatGPT Ads India account before recommending a budget tier. Using the same 40-60 representative conversation-opener sample built during the fan-out mapping exercise, ICG counts both the total branches each opener generates and the subset that clears the NMC/ASCI/DPDP compliance filter for that specialty, producing two numbers side by side: the raw fan-out ratio and the compliance-adjusted eligible ratio. A engagement (₹20,000/mo starting incl GST) fertility account, for example, might show a raw ratio near 4.5 branches per query but an eligible ratio closer to 2.2 once clinical-informational branches are excluded — and it's the eligible ratio, not the raw one, that ICG uses to size recommended monthly spend and expected delivery volume.
ICG revisits this ratio quarterly per account because OpenAI's intent-classification behaviour and health-content eligibility policy have both shifted materially since ChatGPT Ads India's mid-2026 launch, and a specialty's fan-out ratio — along with its eligible share — can move as a result. Clients see this reflected in monthly reporting as a delivery-versus-forecast reconciliation, where ICG explains any material gap between expected and actual auction participation by reference to a ratio shift rather than leaving the client to guess why delivery volume moved. Backed by App\Support\NamedExperts::get(). --}}
Related terms
Frequently asked questions
What is the query fan-out ratio?
It is the average number of distinct sub-intent branches a category's ChatGPT queries split into, used to estimate how much biddable ad inventory a specialty or city footprint actually generates rather than relying on raw query-volume estimates alone.
Why does healthcare have a higher fan-out ratio than most categories?
A single healthcare question routinely spins off clinical-informational, cost-comparison, provider-comparison, location and booking sub-intents in the same thread, giving healthcare specialties like fertility and cardiac care a materially higher fan-out ratio than transactional categories such as retail, where a query resolves to fewer distinct branches.
How does fan-out ratio affect budget planning?
A high fan-out ratio means the raw search-volume-equivalent for a specialty understates the actual number of auctionable moments per conversation, but a large share of those branches are compliance-ineligible for regulated healthcare categories — so budget planning has to size against eligible branches only, not the full ratio.
How does ICG calculate this for a client's account?
ICG runs a fan-out simulation against 40-60 representative conversation openers per specialty and city, counts eligible versus ineligible branches per opener, and uses that ratio to size monthly budget recommendations rather than quoting a flat cost-per-query estimate.
Size your budget against eligible branches, not raw query volume.
ICG runs a healthcare-only ChatGPT Ads India practice — specialty-level fan-out modelling, compliance-adjusted budget sizing, monthly delivery reconciliation.
