How a gastroenterology hospital ran Enterprise-tier ChatGPT Ads for liver-transplant second-opinion intent
The situation
Picture a tertiary hospital running a full hepatology and liver-transplant programme — living-donor and deceased-donor pathways, a hepatobiliary surgery unit, and a general gastroenterology outpatient service seeing several times the transplant programme's volume in patients who will never need a transplant at all. The hospital's existing paid presence, like most hospital groups running a single blended Google Ads account, treated every gastroenterology enquiry the same way, whether it came from someone with mild reflux symptoms or a family whose relative had just been told by another hospital that a liver transplant was the only remaining option.
That blending was expensive in a specific way: transplant-stage enquiries are rare relative to general gastroenterology volume but represent by far the highest-value patient relationship the hospital can acquire, given the length and depth of a transplant care pathway compared with a routine gastroenterology consultation. Yet the hospital's paid channels had no mechanism to isolate transplant-stage intent from general hepatology browsing, meaning transplant coordinators were fielding a mix of genuinely transplant-stage enquiries and much earlier-stage questions that hadn't yet reached transplant evaluation, wasting coordinator time that is itself a scarce and expensive resource in any transplant programme.
A second dimension made the account genuinely complex rather than merely large: living-donor evaluation carries a legal and ethical sensitivity that ordinary hospital marketing doesn't touch, since Indian transplant law strictly prohibits anything resembling solicitation or inducement of organ donation, and the hospital's own legal and transplant-ethics committee needed to sign off on any advertising touching the living-donor pathway before ICG could even begin building creative for that bucket. Post-transplant follow-up care — a genuinely separate service line serving patients years into their post-transplant life needing ongoing hepatology management — added a fourth conversation type entirely distinct from any acquisition-stage intent, rounding the account out to four structurally different buckets under one hospital brand.
The ChatGPT Ads campaign structure ICG designed
Enterprise-tier's parallel-bucket architecture mapped directly onto the programme's real shape: a general hepatology second-opinion bucket for patients and families evaluating whether transplant was genuinely the right next step, a living-donor-information bucket handled under the hospital's ethics-committee-approved framework, a deceased-donor-waitlist-information bucket addressing the very different question of what waiting-list evaluation and MELD-score-adjacent process looks like, and a post-transplant follow-up bucket serving the hospital's existing transplant patient base rather than new acquisition at all. Running these as four separate buckets rather than one blended hepatology campaign let each carry its own bid ladder, its own compliance review depth, and its own landing experience.
Bid ladder design in the second-opinion bucket weighted heavily toward conversations referencing an existing diagnosis of cirrhosis, liver failure, or a prior hospital's transplant recommendation — the clearest signal of genuine transplant-stage intent — at bids illustratively 8-10x above baseline exploratory hepatology conversations about general liver-health questions. The living-donor bucket ran on a structurally different logic: rather than bidding aggressively to acquire volume, its purpose was informational and trust-building, since actual living-donor candidacy isn't something advertising should be driving toward in any direct-response sense, and bids stayed conservative throughout, reflecting the bucket's role as an ethics-conscious information resource rather than a lead-generation engine.
Landing infrastructure ran four matched pages: a second-opinion page detailing the hospital's transplant evaluation process and multidisciplinary team, a living-donor information page built and legally reviewed specifically to describe the evaluation process and eligibility criteria without any language that could be read as encouraging or soliciting donation, a waitlist-and-deceased-donor-process page explaining India's organ allocation system in plain terms, and a post-transplant follow-up page directing existing patients to the hospital's dedicated post-transplant clinic rather than the general enquiry funnel. Each page's enquiry form was tuned to its bucket — the second-opinion form captured current diagnosis and prior hospital, the living-donor form deliberately avoided collecting anything that could resemble a donor-recruitment database, capturing only a request for informational consultation rather than any donor-matching data.
Reporting ran four independent lines feeding one Enterprise dashboard, letting the hospital's transplant coordination team see second-opinion and waitlist volume separately from the hospital's marketing leadership view of the programme's overall paid performance, with the living-donor bucket's reporting reviewed jointly by marketing and the hospital's transplant ethics committee on a monthly basis given its sensitivity.
The compliance discipline
The living-donor bucket carried the account's heaviest compliance weight by a wide margin. Indian law prohibits any commercial inducement or solicitation connected to organ donation, and every ad variant and every line of landing-page copy in that bucket was reviewed against that restriction specifically, not merely against general healthcare-advertising standards — language stayed strictly informational, describing the hospital's evaluation and support process for someone who had already independently decided to explore living donation, never language that could be read as recruiting or encouraging a donation decision. The hospital's own transplant ethics committee reviewed and approved every variant in this bucket before launch, a review layer ICG built into the campaign timeline from the outset rather than treating as an afterthought.
NMC Section 6's prohibition on outcome guarantees applied with particular force across the second-opinion and waitlist buckets, where no copy referenced transplant success rates, graft survival statistics, or waitlist timing in any way that implied a guaranteed or predictable outcome — a discipline that mattered because transplant outcomes and waitlist timing genuinely vary enough, patient to patient, that any implied promise would be both non-compliant and, more importantly, misleading to a family making one of the most consequential medical decisions of their lives. ASCI Chapter III's restriction on fear-based advertising was applied with the same tone-sensitivity ICG uses across all serious-diagnosis specialties, screening not just for factual accuracy but for whether copy read as manufacturing urgency around an already urgent, frightening decision.
DPDP 2023 governed all four enquiry forms, with the second-opinion and waitlist forms requiring explicit consent language for the sensitive diagnosis and treatment-history data they collected, and the living-donor form's consent language reviewed jointly with the hospital's legal team given the additional sensitivity of any data touching a potential donor's health information. The post-transplant follow-up bucket, though lowest-risk of the four from a compliance-claims perspective, still required DPDP-compliant handling of ongoing treatment data for existing patients re-engaging through the paid channel.
The 90-day outcome pattern
In the pattern ICG has observed on comparable Enterprise-tier transplant-programme accounts, the second-opinion bucket ramped first and fastest, since it carried both the highest volume and the clearest intent signal of the four buckets. Illustrative combined conversation-completion events across all four buckets: roughly 25-35 a week in month one, climbing to an illustrative 70-90 a week by month three as the waitlist and living-donor buckets, both lower-volume by design, built out their own smaller but steady conversation streams.
The second-opinion bucket converted to a booked consultation at illustratively 40-50% within 7 days for conversations that had referenced an existing cirrhosis or liver-failure diagnosis, meaningfully higher than the hospital's blended historical conversion rate across its general Google Ads hepatology presence, consistent with the expected pattern that isolating genuine transplant-stage intent from general hepatology browsing produces a higher-quality, faster-converting lead pool even at lower absolute volume. The living-donor bucket, deliberately run for information and trust rather than volume, generated a small but consistent stream of informational-consultation requests — illustratively 4-6 a week by month three — each one representing exactly the kind of carefully-considered, already-decided enquiry the bucket was designed to serve rather than to generate through persuasion.
The post-transplant follow-up bucket, though not an acquisition play in the traditional sense, produced a downstream benefit the hospital's marketing team hadn't fully anticipated: illustratively 60-70% of follow-up-bucket conversions came from patients who had moved cities or changed contact details since their transplant, meaning the bucket functioned as a re-engagement channel recovering patients who might otherwise have drifted to a different hospital for routine post-transplant management. Blended cost per qualified second-opinion lead across the account ran illustratively 20-30% below the hospital's blended Google Ads cost per hepatology lead over the same quarter, with GA4's AI Assistant channel showing a key-event rate consistent with ICG's broader healthcare book, and the transplant coordination team reported a noticeably lower share of clearly-premature enquiries reaching their desk compared with the pre-engagement blended-funnel period.
What we'd do differently next time
Launching all four buckets simultaneously meant the living-donor bucket's ethics-committee review process, which needed more back-and-forth than any other bucket's compliance review, held up the account's overall launch timeline by roughly two weeks even though the second-opinion and waitlist buckets were ready to go live immediately. Launching the second-opinion and waitlist buckets first, with living-donor and post-transplant follow-up following once their respective review processes completed, would have gotten the highest-volume, most time-sensitive buckets live faster without waiting on the slowest review track.
The post-transplant follow-up bucket's re-engagement value wasn't part of the original brief, and it took until month two for the hospital's own patient-relationship team to notice and act on the pattern of address and contact-detail updates coming through that bucket. Building an explicit hand-off process from the follow-up bucket to the hospital's patient-records team from day one, rather than discovering the pattern informally, would have let the hospital capture that operational benefit sooner.
Finally, the waitlist-information bucket's landing page, while accurate, ran more like a general educational page than a genuinely differentiated one, and the hospital's own transplant coordinators felt it underexplained what made the hospital's specific allocation-support process worth choosing over another transplant centre. A more specific, hospital-differentiated version of that page, built with more direct coordinator input from the outset, would likely have converted waitlist-stage conversations at a higher rate.
How this maps to your own transplant or hepatology programme
If your hospital runs a transplant programme alongside a much higher-volume general specialty service — true of most transplant centres, whether liver, kidney, or heart — this Enterprise-tier structure of isolating transplant-stage intent into its own bucket, running any donor-related messaging under a dedicated ethics-and-legal review track, and treating post-transplant follow-up as its own distinct conversation type rather than folding it into acquisition, is worth evaluating even before your programme's volume justifies full Enterprise budget.
The core transferable lesson is that transplant-adjacent advertising carries legal and ethical weight beyond ordinary healthcare-advertising compliance, specifically around anything touching organ donation, and that building a dedicated review track for that content from the outset — rather than applying general healthcare compliance review and hoping it's sufficient — is the only responsible way to run this category of campaign.
