Google would love to run your bids from day one. Every new account gets the pitch: turn on Smart Bidding, set a goal, let the AI handle it. For a practice with no bookings on record, that's handing the car keys to a driver who has never seen a map. The AI is good. It just has nothing to learn from yet.
So the method this guide teaches starts the other way around. You set the bids by hand for the first few months, not because manual bidding is better forever, but because it's how the account learns what a new patient actually costs. Once that number is real, it goes to Google as the target, and the AI finally has something worth chasing.
Manual Bids Are Tuition. The Lesson Is Your Real Cost per Patient.
Manual CPC is the simplest bidding there is. You set the most you're willing to pay for one click, and Google never charges more than that. Google's help center describes two ways to apply it: one default bid for a whole ad group, or separate bids on the keywords inside it when you know one search books better than the rest.
In the first campaign, the one we call Isolation at BrandRocket, bids live at the ad group level. Each ad group is one service with its own page, as Chapter 5 built them, so a bid on the "orthopedic surgeon" group is a bid on one kind of patient. If one keyword in that group starts booking far better than its neighbors, it gets its own bid.
What you're really buying in these months isn't clicks. It's information. Every click that turns into a booked patient, and every click that doesn't, adds to the one number the rest of this chapter depends on: what it costs this account, in this market, to put a new patient on the schedule for this service.
Chapter 3 priced a new patient in each specialty against what those visits actually pay. That was the math on paper. Isolation finds out what the market actually charges.
The Launch Bid Is High on Purpose and Comes Down as Quality Score Builds
A brand-new account starts at the back of the line. Quality Score leans partly on how often Google expects your ad to be clicked, and a practice that just opened its account has no click history to lean on. The way through is to bid high enough at launch to earn a spot on the page while the account builds a record.
Keyword Planner gives you the opening number. Its top-of-page range shows what advertisers have historically bid to reach the top of the page, and as Chapter 5 warned, it's an estimate you should read as about 30% either way. For "orthopedic near me," our example market, San Diego, shows $2.37 to $7.41. A launch bid near the upper edge of that range is on purpose, not a mistake.
Then it comes down. As the ads collect clicks and the landing pages prove they answer the search, Quality Score climbs, and the same position costs less. Every week or two in the early months, check the bids against the page position they're buying, and lower the ones that are winning comfortably. Nobody gets a medal for keeping a launch bid in place six months later.
The Top of the Page Tops Out Near $4 for a Cardiologist and $38 for a Therapist
Here's where healthcare gets strange: the price of the top of the page depends almost entirely on the specialty. This is the top of Keyword Planner's price range for the main "near me" search in nine specialties, from our San Diego data:
- "cardiologist near me," up to $3.89
- "podiatrist near me," up to $5.00
- "obgyn near me," up to $5.31
- "orthopedic near me," up to $7.41
- "allergist near me," up to $8.78
- "physical therapy near me," up to $18.86
- "fertility places near me," up to $24.16
- "psychiatry near me," up to $25.63
- "therapist near me," up to $38.38
Most of the specialties we checked top out under $9. Physical therapy, fertility, psychiatry and therapy run roughly two to four times that. Therapy is also where the big national players advertise. Among the advertisers we checked for this guide, a national telehealth therapy network had at least 15 ads on record in Google's Ads Transparency Center, and so did a national therapist directory.
For a multi-specialty practice, this is the practical point: a click price that's normal in one service line would be reckless in another. Give each service line its own campaign so its bids, its budget and, later, its target stand on their own numbers. Chapter 3 made the same case for budgets. It applies twice over for bids.
Top Impression Share Shows Whether a Bid Is Buying the Top of the Page
A bid is only a guess until the account shows you what it bought. Before the first ad runs, add three columns to the keyword view. Together they turn every bid into a report card:
- Search top impression share: the share of eligible top-of-page showings, above the organic results, that you actually got. On a practice's most important searches, the method this guide teaches aims for 75% to 90%.
- Search absolute top impression share: the share of the time your ad sat in the very first spot. Nice to have on your top two or three searches, and not worth chasing everywhere.
- Search lost impression share (rank): how often you missed out because your Ad Rank came up short. Google also reports lost impression share from budget, but only at the campaign level.
Those two losses mean different things. A shortfall from rank means the bid is too low, or the ad and page haven't earned their Quality Score yet. A shortfall from budget means the money ran out before the day did, and raising a bid only makes that happen sooner. Quality Score itself sits at the keyword level; check it next to the bids.
Auction Insights Tells You Which Fights Aren't Worth Winning
Some auctions aren't worth winning at any price. For a practice, the auction insights report is the waiting room seen from the other side: every rival practice, network and directory whose ads land on your searches, each scored on six measures, including how often they shared the page with you and how often they finished above you.
Read it with one question in mind: who's sitting next to me, and can I afford to outbid them? A local therapy practice that sees national telehealth networks and directories dominating "therapist near me" has a choice. It can pay whatever they're paying, or it can step back from the searches where they pile up and keep bidding where it wins: a neighborhood, a specific kind of therapy, a problem the networks don't treat. Walking away from one expensive search to win three affordable ones isn't retreat. It's bidding with your eyes open.
The Cost per Booking Your Exact-Match Campaign Earned Becomes Google's Target
This is where the manual months pay off. Once a service line is booking about 30 patients a month, it's ready for Exploration: the broad-match campaign from Chapter 6, run on Maximize Conversions with a target cost per booking. In June 2026, Google began shortening that setting's name to just "Target CPA," so older and newer screens may label it differently for a while.
The target comes from Isolation's real numbers, not from a guess. Google's own guidance lines up with this. If the campaign already has bookings on record, Google suggests a starting target built from its average cost per conversion in the past month, and it advises measuring results over at least 30 conversions. That's the same 30-booking threshold the method this guide teaches uses to switch Exploration on.
Two warnings from Google's help center are worth taking seriously. First, a target set too low can cost you bookings, because the system passes on clicks that might have converted. Starting at the real average and lowering it slowly beats starting at a wish. Second, Google lets you set separate targets for individual ad groups, but doesn't recommend it. For a multi-specialty practice, that's one more reason each specialty runs as its own campaign: an orthopedic booking and a therapy booking don't cost the same, and one blended target would overpay for one and starve the other.
A Cost per Booking Is Only as Honest as the Calls Behind It
Before any number goes to Google as a target, check what it's made of. In most practices, calls are a big share of the bookings. Chapter 1 set phone calls of 30 seconds or longer to count as conversions, and that's the right floor: a patient who leaves a voicemail after hours is often a patient. But a 30-second call can also be an existing patient moving an appointment, someone asking whether you take their insurance, or a wrong number that took a minute to untangle.
If those calls are counted as bookings, the cost per booking looks better than it is. Hand that number to Google as a target, and the system gets very good at finding more of the calls that don't become patients.
The fix is unglamorous. Every week in the early months, compare the calls Google counted against what the front desk actually scheduled. If your call tracking records calls, listen to a sample, and keep those recordings inside the same privacy rules you apply to any patient information. When the real count and the reported count disagree, trust the front desk and adjust the target before you hand it over.
Our specialty guides cover bidding in their own fields. The dermatology guide covers setting bids by condition. The plastic surgery guide covers what a click costs by procedure. The optometry guide covers holding the top of the page through the January rush.
Next: Ads Written in the Words Patients Actually Use
The bids decide where the ad shows up. The ad itself decides whether anyone clicks. Chapter 8 builds ad copy from your own pages and the words patients already use about you.
Catch up on Chapter 6 on exact match and the long tail, or see every chapter on the guide home for Google Ads for Healthcare. When a practice hires us for Google Ads management, every new account starts on manual bids and earns its way to a target.
