Your market is smaller than the keyword tool says
Depression search volume is enormous. The share of it that meets insurance criteria for TMS is small. Most plans require documented failure of two or more adequate antidepressant trials, often plus a trial of psychotherapy, before they will authorize a course.
Bidding on depression treatment brings you calls from people who have never tried a medication, and your staff spends the consult explaining that they are not a candidate yet. That is an expensive way to fill a calendar.
The terms that work name the situation: treatment resistant depression, depression when medication does not work, TMS therapy near me, and the device brand names patients search after a psychiatrist mentions them.
Insurance is the objection, so make it the campaign
The first question a prospective TMS patient asks is whether it is covered. Medicare and most commercial plans cover TMS for major depressive disorder that meets criteria. Without coverage the patient pays out of pocket for a six-week course of daily sessions.
Ad groups built around payer names, with a landing page explaining that payer's criteria, convert at a rate general depression campaigns cannot approach. It is also cheaper inventory, because the competitors are all bidding on the condition.
The referral search is the one most clinics miss
A large share of TMS patients hear about it from their psychiatrist and then search. They search the modality, the device brand their doctor named, and your clinic's name.
Brand and modality campaigns are inexpensive and they protect a referral you already earned. Losing your own clinic name search to a competitor bidding on it, or to a national provider, throws away referral relationships that took years to build.
The restrictions you are working inside
Google prohibits personalized advertising based on inferred mental health status, which removes health-based remarketing and the similar audiences built from it. Campaigns run on keyword intent, geography, device, and schedule.
Ad copy also cannot promise outcomes. TMS response and remission figures come from published trials, and putting them in a 90-character headline is a policy risk and a clinical one. Save the evidence for the landing page, where it can be sourced.
Count consults, not clicks
The path is a call, a screening for candidacy, an insurance verification, then a course of daily treatments over roughly six weeks. That is a high-value patient and a long enough sequence that click metrics tell you nothing useful.
Instrument the phone with dynamic number insertion by campaign and keyword, then have the intake team mark which calls became verified candidates. The report reads cost per call, cost per qualified candidate, and where the CRM supports it, cost per started course.