Oncology centres ChatGPT Ads India — capture patient intent inside the conversational search that pre-consultation happens on

Conversational-search behaviour specific to oncology centres patients in 2026

The oncology patient journey is unusually well suited to conversational search because the entry point is almost always a moment of confusion rather than a moment of clarity. A patient receives a histopathology report full of terms they have never seen before, or a radiologist's staging note, or a first oncologist's proposed protocol, and their immediate instinct is not to type three keywords into a search bar — it is to paste the report or describe it in full sentences and ask an assistant to explain it. This is precisely the interaction pattern ChatGPT and similar tools are built for, and it is precisely the moment a hospital's paid search strategy has historically had no way to reach.

What follows that first explanatory exchange is a multi-turn conversation, not a single query. A caregiver researching a parent's lung cancer diagnosis will ask what stage 3 means, then ask what immunotherapy is and whether it applies to this case, then ask what the treatment typically costs in India, then ask how to find a good oncology team nearby. Each of those turns is a distinct intent signal, and a hospital that is only bidding on the keyword equivalent of the last turn is missing the four conversations that built the trust to get there. ChatGPT Ads bid across this entire arc, which is structurally different from a keyword auction that only fires at the moment of search.

Caregivers, not only patients, drive a large share of this volume. In our data across oncology-adjacent healthcare clients, a substantial portion of conversational research sessions around a cancer diagnosis are conducted by an adult child or spouse rather than the patient themselves — often at night, often across multiple sessions over several days, and often comparing what one hospital's assistant-cited page says against another's. A landing page that reads as generic marketing copy loses this comparison; one that reads as a clear, factual explanation of protocols, team credentials and next steps wins it.

The second-opinion pattern is the clearest evidence of this shift. Patients who already have a diagnosis and a proposed treatment plan increasingly use conversational search to sense-check it before committing — asking whether the proposed chemotherapy regimen is standard for their cancer type and stage, or whether a competing centre's immunotherapy protocol is worth exploring. This is a caregiver and patient population that has already accepted the need for treatment; the only decision left is which centre and which oncology team, and that is a decision a well-positioned conversational ad and a well-built landing page can genuinely influence.

Intent bucket map — stages, example queries and CPC bands

Oncology conversational search does not behave as a single funnel; it behaves as five distinct intent buckets, each with a different cost profile in the ChatGPT Ads auction, because the auction prices conversation-stage rather than keyword volume. Structuring an account around these buckets — rather than around a flat list of oncology keywords — is what separates a campaign that spends efficiently from one that burns budget on early-stage exploration traffic that was never going to convert this quarter.

The gap between stage 1 and stage 5 pricing — roughly fifteen to twenty times — is the single most important planning number for an oncology CMO evaluating this channel. It means a naive account structure that lets the auction find its own level, without deliberate budget allocation across stages, will drift the majority of spend toward stage 1 and 2 queries simply because they are cheaper and far more numerous. That volume matters for brand presence and long-term nurture, but it will not fill an oncology OPD calendar on its own. ICG's account structure deliberately reserves 55 to 65% of oncology budget for stage 3 through 5, with the remaining share funding stage 1 and 2 for pipeline building and remarketing eligibility.

Radiation oncology and surgical oncology queries cluster more heavily in stages 3 through 5 than medical oncology does, because a patient asking about stereotactic radiosurgery or tumour resection has typically already passed through the comprehension and options-research stages with their diagnosing physician. This means a radiation or surgical oncology campaign can often run a leaner, more expensive-per-click but higher-converting structure, while a medical oncology campaign — which owns the earliest and broadest part of the funnel, including chemotherapy and immunotherapy exploration — needs deeper stage 1 and 2 coverage to build the audience that later converts on stage 4 and 5 queries.

Compliance overlays for oncology conversational advertising

Oncology sits at the highest compliance-risk end of India's healthcare advertising spectrum, and every asset ICG builds for this vertical is checked against four overlays before it goes live. NMC Section 6 prohibits solicitation of patients and restricts advertising that implies guaranteed outcomes — a real risk in oncology copy, where the temptation to reference survival rates or treatment success percentages is strong and the regulatory line is unforgiving. ICG's copy discipline for oncology never states outcome statistics as a promise; it states protocol availability, team credentials and facility capability as facts, and leaves outcome discussion to the clinical consultation.

ASCI Chapter III governs substantiation of health claims, and oncology claims are scrutinised more closely than most categories because the stakes for a misled patient are severe. A claim that a centre offers "advanced immunotherapy" must be substantiable — the specific therapies, the specific approval status, the specific patient-eligibility criteria — rather than used as a generic marketing adjective. ICG's compliance review checks every therapy-name reference in oncology ad copy and landing pages against what the centre can actually document before publication.

DPDP Act 2023 governs how patient data collected through an oncology landing page's enquiry form is handled, and this carries particular weight in oncology because the data being submitted is not just contact information — it is often a description of a cancer diagnosis, a stage, a treatment history. ICG builds oncology landing pages with explicit consent language at the point of form submission and ensures the data-handling disclosure matches what the hospital's actual CRM and data-retention practice is, not a generic template.

The oncology-specific overlay beyond these three general rules is around cost and access framing. Because cancer treatment cost is one of the single highest-anxiety, highest-search-volume topics in this category, ad copy and landing pages that reference cost must do so as factual package information — inclusions, typical ranges, what varies by stage and protocol — never as a comparative claim against another named centre, and never framed in a way that could be read as steering a treatment decision toward a cheaper option regardless of clinical appropriateness. This is not only a compliance requirement; it is also what makes the landing page trustworthy enough for a frightened caregiver to act on.

The landing-page pattern that converts oncology ChatGPT Ads traffic

A conversational assistant does not send a click the way a search engine does — it cites a source and, when the user follows through, delivers them to a page that the assistant has already effectively summarised on the user's behalf. This means the landing page's job is not to persuade from a cold start; it is to confirm and extend what the assistant has already told the user, quickly and without friction. A dense, promotional hospital department page — built for a scrolling desktop visitor with time to browse — routinely underperforms here, because it forces the user to re-derive information the assistant already gave them.

ICG's oncology landing-page pattern opens with a short, factual answer block addressing the specific query cluster the ad was served against — for example, a cost-and-package page opens with a plain-language range and inclusion list for the relevant cancer type and treatment modality, stated in the first 100 words, before any hospital-credentialing content appears. This mirrors how the assistant itself would summarise the page if asked to cite it again, which matters because these pages are frequently re-cited in follow-up conversation turns after the first visit.

Below the answer block, the page carries oncology team credentials — named consultants, sub-speciality, years of oncology-specific experience, tumour board structure — because second-opinion and ready-to-consult traffic converts on team credibility more than on facility photography. A short, structured FAQ section addressing the two or three most common follow-up questions for that query cluster (does this apply to my stage, is a second opinion free, how soon can I get an appointment) closes the page, again in the short, scannable format that both human readers and conversational assistants parse well.

The conversion action itself is kept to a single low-friction step — a callback request form with three fields, not a multi-step patient-intake form — because oncology enquiries at this stage of the journey are emotionally loaded, and every additional form field is a point where a caregiver abandons rather than completes. ICG tracks conversation-completion and form-submission events separately in GA4 under the AI Assistant channel, which lets a hospital see not just how many people arrived from ChatGPT Ads but how many completed the intent journey the ad was built to serve.

90-day expected outcomes

The first 30 days are account-build: intent-bucket structuring, compliance-cleared ad copy across all five stages, and the machine-scannable landing pages described above, built or adapted for each query cluster the centre wants to compete in. This phase produces limited lead volume by design, because the auction needs conversation history to price stages correctly and the compliance review needs to run before anything scales.

Days 30 to 60 typically show rising engagement in stage 1 through 3 buckets first, since these have higher volume and lower CPCs, followed by the first meaningful cluster of stage 4 and 5 conversions as remarketing eligibility builds from the earlier-stage audience. This is also when the first clean read of the AI Assistant channel's conversion rate against a hospital's own baseline becomes available in GA4, letting budget be reallocated toward whichever oncology sub-speciality is producing the strongest cost-per-qualified-lead.

By day 90, a well-run oncology ChatGPT Ads account should show a stable, repeatable contribution to qualified enquiry volume, with cost-per-qualified-lead data solid enough to justify the next quarter's budget decision — whether that means holding tier, scaling up, or narrowing focus to the highest-performing sub-speciality. Given the 90 to 180-day whitespace window before more hospital groups enter this channel in India, centres that reach a stable 90-day baseline now are positioned to hold lower acquisition costs through the period when the auction is expected to tighten.