How a dental chain in tier-2 cities graduated from Starter to Scale tier over 8 months
The situation
Picture a dental chain running six branches across three tier-2 cities in North and Central India — Meerut, Bareilly and Gwalior, in the shape of this scenario, though the specific cities matter less than the pattern they represent. The chain had built a reasonable local reputation over a decade, mostly through word-of-mouth and a modest Google Ads presence that covered general dental keywords without much branch-level targeting. Its owner-dentist ran marketing decisions personally, which meant budget discipline was tight and every rupee spent had to show a traceable path to a booked appointment.
The problem wasn't lack of patients — it was inconsistent lead quality across branches. Two of the six locations, both in newer commercial districts, consistently underperformed on Google Ads despite similar spend to the older, better-established branches. The chain's marketing coordinator suspected the issue was that Google Ads keyword auctions in tier-2 markets were increasingly crowded with aggregators and directory listings bidding up generic dental terms, leaving little room for a chain of this size to compete on cost per click alone.
ChatGPT Ads entered the conversation not as a strategic bet but almost by accident — the marketing coordinator noticed a competitor's clinic name surfacing in a screenshot a patient shared, apparently from a ChatGPT conversation about finding a dentist nearby. That was enough to prompt a scoping call. The chain's initial ask was modest and specific: could a small, controlled budget test whether conversational search produced better-qualified leads than the increasingly saturated Google Ads auction, without committing to a spend level the owner-dentist wasn't ready to justify yet. That modesty is exactly what Starter tier is built for — a controlled, single-bucket test rather than a full multi-branch rollout from day one.
The ChatGPT Ads campaign structure ICG designed
ICG's Starter-tier build for this chain ran as a single conversation bucket covering all six branches, rather than six separate campaigns. At this budget level and this stage of data-gathering, splitting the account six ways would have starved every branch of enough conversation volume to reach a stable bid ladder. Instead, the single bucket used a locality-aware bid modifier — conversations that named or implied one of the chain's three cities bid slightly higher than generic dental-concern conversations, on the logic that geographic specificity is itself a weak signal of intent.
The bid ladder itself followed the standard conversation-stage structure: early-exploration conversations (a user asking general questions about a dental concern, tooth sensitivity, or whether they need to see a dentist at all) bid at the floor of the range, while late-intent conversations (a user asking about booking a specific procedure, comparing costs, or asking which of the chain's branches was nearest) bid 6-9x higher. Given Starter tier's budget ceiling, ICG capped the account's daily spend on the highest-intent tier deliberately conservatively in month one, prioritising enough exploration-stage volume to build a real dataset over chasing a handful of expensive late-intent conversions immediately.
Landing infrastructure at Starter tier was intentionally lean — one machine-scannable landing page covering all six branches, with a branch-selector as the primary interaction rather than six separate pages. This kept production cost low while the account was still proving itself, and it meant every conversation-attributed lead flowed into one enquiry form the owner-dentist's front-desk team could manage without new process overhead. Reporting ran monthly, in a single dashboard, deliberately simple — completed conversations, cost per qualified lead, and a manual reconciliation against the front desk's appointment book, because at Starter-tier volumes automated CRM integration wasn't yet worth building.
By month three, conversation volume and cost-per-lead data were stable enough to show a clear pattern: the two underperforming branches on Google Ads were actually the chain's strongest performers on ChatGPT Ads, plausibly because their newer commercial-district locations matched the kind of specific, comparison-oriented conversations the assistant channel was surfacing. That data point was the trigger for the Growth-tier upgrade — enough evidence to justify widening the bid ladder and building branch-specific landing pages for those two locations rather than continuing to run all six through one shared page.
The compliance discipline
Dental advertising sits under the same baseline as the rest of India's healthcare advertising regime — NMC Section 6 and ASCI Chapter III — but carries its own recurring pressure points that ICG's compliance review specifically screened for in this account. Implant and aligner procedures are the two categories where dental marketing most often drifts into outcome-guarantee language, and every ad variant referencing either procedure was checked to ensure no phrasing implied a guaranteed cosmetic or functional result. "Painless" and "same-day" claims, common in dental marketing generally, were screened out unless the chain could substantiate them procedure-by-procedure, which in most cases it couldn't do to the standard ICG's review required.
Before-after framing was the second recurring issue. The chain's existing marketing materials included several before-after style descriptions of cosmetic dentistry outcomes, carried over from its pre-ChatGPT-Ads Google Ads copy. ASCI Chapter III treats these as a compliance risk when framed as a promise rather than an illustrative example, so ICG's copy review rewrote every instance to describe procedures factually — what a treatment involves, typical recovery expectations sourced from general clinical literature rather than the chain's own claimed results — without implying a specific patient's outcome as representative.
Comparative language against other dental providers in the same tier-2 cities was the third area of discipline. Given that the chain's initial interest in ChatGPT Ads was triggered by seeing a competitor surface in conversational search, there was an early instinct on the client side to write copy that positioned the chain against named local competitors. ICG declined to run any comparative claims naming or implying inferiority of specific competing clinics — both because ASCI Chapter III restricts this and because category-level positioning (why choose a multi-branch chain with consistent protocols over an independent single-dentist practice) achieves the same commercial goal without the compliance exposure. DPDP 2023 governed the enquiry form across all branches from day one, with explicit consent language and no default data-sharing between the chain and any third party.
The 90-day outcome pattern
Illustrative numbers for this hypothetical: the Starter-tier bucket produced roughly 15-20 completed conversations a week by the end of month one, rising to around 35-40 a week by month three as the shared bid ladder stabilised. Cost per qualified lead started high in week one — the data-gathering period any new ChatGPT Ads account runs through — and settled by month two into a range the owner-dentist judged comparable to, and in the two underperforming branches' catchments meaningfully better than, the chain's existing Google Ads cost per lead.
The conversion-to-appointment pattern is what actually drove the tier upgrade decision, not raw lead volume. Across the six-branch shared bucket, roughly 40-45% of conversation-attributed leads converted to a booked appointment within 10 days — a rate the chain's front desk described as noticeably cleaner than its Google Ads leads, where a larger share turned out to be price-shopping calls that never booked. In GA4, this traffic aggregated under the AI Assistant channel, consistent with the elevated key-event rate ICG tracks across its healthcare book generally, well above the chain's organic and direct traffic baselines.
The Growth-tier upgrade at month three widened the bid ladder for the two branches showing the strongest performance and added dedicated landing pages for each, while keeping the remaining four branches on the shared Starter-style page for another data cycle. This selective upgrade — rather than a wholesale tier jump for all six branches simultaneously — let the chain justify the increased spend incrementally, with each branch's own numbers making the case for its next round of investment.
By month six, four of six branches had shown a stable enough pattern to justify the Scale-tier structure: dedicated landing pages for every branch, procedure-category buckets (implants, aligners, general dentistry, cosmetic) split out from the single locality-modified bucket, and monthly reporting reconciled against the chain's now-digitised appointment system rather than a manual front-desk log. By month eight, the chain's blended cost per qualified lead across all six branches on ChatGPT Ads sat in the range ICG has observed repeatedly for mature healthcare accounts — meaningfully below its blended Google Ads cost per lead, with the caveat the chain's own team raised unprompted: this reflected a still-uncrowded channel in tier-2 markets specifically, a window ICG expects to narrow as more dental and healthcare advertisers discover conversational search.
The procedure-category split introduced at Scale tier revealed a pattern the branch-level view alone hadn't surfaced. Implants and aligners, run as their own buckets from month six onward, showed sharply different conversation-stage profiles from each other despite both being "premium" procedure categories in the chain's internal pricing structure. Implant conversations skewed heavily late-intent — patients researching implants had usually already ruled out alternatives and were comparing providers on specifics like material, warranty and aftercare, which meant the implant bucket ran a higher average cost per conversation but also the chain's strongest conversion-to-appointment rate of any category by month eight. Aligner conversations, by contrast, stayed in exploration mode far longer, with many users still comparing aligners against traditional braces or weighing whether treatment was worth the cost at all — useful data that let the chain's front desk prepare a more tailored first-call script for aligner enquiries specifically, addressing the cost-comparison question before the patient had to ask it.
General dentistry and cosmetic dentistry, the two remaining Scale-tier buckets, produced the chain's highest raw conversation volume but the lowest average value per booked appointment, consistent with how these categories function as entry points into the practice rather than high-ticket decision points on their own. The chain's owner-dentist noted that several patients who first engaged through a general-dentistry conversation later returned as implant or aligner patients months afterward — a downstream pattern ChatGPT Ads' conversation-completion attribution didn't capture directly, since it measured the first conversion event rather than lifetime patient value, but one the chain's own CRM eventually surfaced once appointment history was cross-referenced against original lead source over a longer window.
What we'd do differently next time
The single biggest inefficiency in this account's growth path was building branch-specific landing pages reactively, only after the Growth-tier trigger fired, rather than having a lightweight branch-page template ready to deploy the moment data justified it. That gap cost roughly two to three weeks of production lag at each tier transition, weeks where the account could have been capturing higher-quality traffic through dedicated pages but was still running on the shared Starter-tier landing page.
The locality bid modifier in the Starter-tier bucket, in retrospect, should have been paired with a lighter-weight version of branch-level reporting from month one — not full dedicated dashboards, but a simple breakdown of which branch's catchment each conversation implied. Without that granularity, the signal that two specific branches were outperforming didn't become visible until the month-three data review, when a branch-tagged report from week one would likely have surfaced the same pattern by week six or seven, accelerating the whole upgrade timeline by roughly a month.
Finally, the compliance rewrite of the chain's before-after cosmetic dentistry language happened after the account was already live, because the client's existing marketing materials weren't fully reviewed during onboarding. Front-loading a full compliance audit of all existing client-supplied copy before campaign launch, rather than during the first month of live review, would have caught this earlier and avoided a mid-flight copy revision that briefly paused the cosmetic-dentistry conversation segment.
How this maps to your own dental practice
If you run a single-location dental practice or a small two-to-three branch operation, the Starter-tier structure described here — one shared conversation bucket, one lean landing page, simple monthly reporting — is likely the right entry point, not the multi-branch, procedure-split Scale-tier structure this chain grew into. The tier upgrade path matters more than the tier you start at: what justified each step up in this scenario was data showing a specific branch or procedure category had enough volume and conversion strength to earn its own dedicated infrastructure, not a fixed timeline or an arbitrary ambition to reach the top tier.
Whatever tier fits your practice, the compliance baseline is identical — NMC Section 6 and ASCI Chapter III govern every dental ad, with particular scrutiny on implant and cosmetic-procedure outcome claims regardless of how small or large the account is.
