ChatGPT ads for dental practices, run as a channel.

As a measured test alongside your existing channels, usually yes; as a replacement for search, not yet. The audience is unarguable at a billion weekly users with around 20% showing commercial intent, and advertiser numbers are climbing fast, which means today's auction is quieter than next year's. We size the test so a disappointing result costs you a modest number rather than a quarter's marketing budget.

Through OpenAI's ads manager on a cost-per-click basis, with conversion bidding available once the web pixel is installed and matching properly. There is no lookalike targeting, no interest targeting and no video placement, so the levers are fewer than you are used to in Google or Meta. That is genuinely fine for a single-site clinic: the effort moves to the questions you buy, the copy that answers them, and the page behind the click.

With the web pixel on your site, tracked phone numbers on the pages the ads point at, and a booking flow that records its source. We verify each event against a real test booking rather than trusting a dashboard, and we reconcile against your practice management system monthly. Where attribution genuinely cannot be proven, we say so on the report instead of assigning the patient to whichever channel flatters us.

Some overlap is inevitable, because a patient who researched in an assistant may still search your practice name before booking. That is why we run the two from one plan and report them side by side, watching total cost per new patient rather than each channel's own scorecard. In practice the assistant tends to capture earlier, longer research conversations, while search keeps the urgent same-day demand.

The same rules apply here as everywhere else: no guaranteed outcomes, no misleading price claims, no implying a treatment is risk-free, and care with before-and-after imagery depending on your regulator. We write to answer the question a patient asked and to invite a consultation, which is both compliant and more persuasive than a superlative. Where a claim is borderline we route it past your clinical lead before it runs.

Yes, and it makes the paid placement work harder. Assistants lean on public evidence: your reviews, your profile, the health content attributable to a named dentist. 97% of consumers read reviews for local businesses, and AI tools now sit alongside Google as a source of local recommendations, so review velocity and accurate practice information are part of this project rather than a separate SEO conversation.

Enough to produce readable data inside a quarter, which for a single-site practice is usually a small four-figure monthly amount rather than a token few hundred. Ad delivery is still uneven while the platform ramps, so a thin budget produces noise you cannot act on. We agree the cap in writing, review it monthly, and will recommend stopping rather than quietly rolling the budget forward if the numbers do not hold up.

The ones patients research before they book. Implants, orthodontics and clear aligners, and cosmetic cases all involve long comparison conversations about cost, recovery and alternatives, which is exactly what people bring to an assistant. Emergency dentistry behaves differently and still belongs mostly in search and local. Hygiene sits in between, and we usually let the first eight weeks of data settle that rather than argue about it.

You do. The ads manager account, the pixel, the call tracking and the analytics are set up in your name, and we work inside them as a user. If our engagement ends, nothing is switched off and nothing has to be rebuilt: the campaigns, the history and the documentation stay with the practice. We have no interest in holding an account hostage to keep a client.

It is one plan, not three. The same research that tells us which questions to buy tells our AI search team which answers to earn and our local SEO team which pages and profiles need attention. Reporting shows paid and organic visibility next to each other so nobody claims the same new patient twice, which is the usual outcome when three vendors report separately.

It will, repeatedly — the pilot went from nothing to a hundred million dollars in annualised revenue in under two months, and the product is still moving. We keep the setup simple so it can be rebuilt quickly, note platform changes on your monthly report, and adjust the plan rather than waiting for a quarterly review. Nothing we build depends on a feature that only exists this month.

Yes, and the numbers get more honest when one team holds them. Our dental PPC, paid social and analytics specialists work from the same plan as this channel, so a new patient is attributed once. Hiring us for this test alone is equally fine — everything is documented and left in accounts you own.