Someone with a bad knee doesn't search for "orthopedics." They search for their knee: "knee replacement surgeon," "torn meniscus treatment," "acl surgery." Someone with a stiff shoulder searches for their shoulder. The person with numb fingers at night searches for carpal tunnel.
Orthopedic patients arrive sorted by body part. Most orthopedic ad accounts un-sort them, pouring every search into one "Orthopedics" bucket with one set of ads and one landing page. Google notices, and it charges accordingly.
Google Doesn't Just Sell the Top Spot. It Grades Who Deserves It.
Each search sets off a quick auction that picks which ads appear and in which order. It's tempting to think the biggest bid wins. It doesn't, at least not by itself.
Google names six things that feed Ad Rank, the score behind your position. Your bid is one. The others: the quality of the ad and its landing page, judged live at search time; the minimum bar an ad must clear before it can appear; how crowded that auction is; who is searching, from where, on what device and when; and how much your assets and other ad formats are expected to add. Google doesn't publish a formula that combines them, and anyone who shows you one is guessing.
What matters for an orthopedic practice is the second item. A practice can win a higher spot than a health system bidding more, because its ad and page are the better answer to that search. And position matters: ads near the top of the page get the most clicks, so being the better answer is also how a smaller practice gets seen.
Quality Score Grades Three Things, and Your Bid Isn't One of Them
Each keyword also carries a Quality Score between 1 and 10. It's easy to treat it as the thing to chase. Google says plainly that it isn't part of the auction. It's a diagnostic, a report card that tells you where your ads and pages are weak.
Three grades make it up:
- Expected click-through rate: Google's forecast of how often people who type that keyword will tap your ad. Position and extras are factored out of that forecast, so you can't buy a better grade with a higher bid or more assets.
- Ad relevance: how well the ad's words line up with the search that triggered it.
- Landing page experience: how useful and relevant the page behind the ad is to that searcher.
Each one comes back as above average, average or below average. That's the useful part. A below-average ad relevance tells you the ad doesn't match the search. A below-average landing page experience tells you the page doesn't. You don't fix a Quality Score. You fix whichever of the three it's pointing at.
To see it, open the keywords report in Google Ads and add the Quality Score columns: the overall score plus a column for each of the three parts. A new keyword shows no score until it has enough searches behind it, so give a new account a few weeks before reading anything into the numbers. When you do look, sort by spend, not by score. A low score on a keyword that barely runs doesn't matter. A low score on your most expensive knee keyword is costing you every day.
Patients Search by Body Part. Most Accounts Are Built by Department.
Ad relevance is the grade you can move fastest, and in orthopedics it's decided by how you sort your keywords.
In our example market, Nashville, the keyword ideas we pulled from twelve orthopedic websites fall cleanly by body part: knee, ACL and meniscus searches led with 1,576 different phrasings; hip had 332; spine, back and neck 222; hand and wrist 142; shoulder 127; and foot and ankle 88. Several of those lists came mostly from the one specialist site that seeded them, so treat the counts as what one market's sites produced, not a national census. The shape is what matters, and it holds nationally too: the largest procedure searches are body-part searches, like "knee replacement" at 33,100 a month and "acl tear" at 74,000.
So build the account the same way. Instead of one "Orthopedics" ad group holding knee, hip, spine and hand searches together, give each body part, and then each procedure within it, its own ad group of about five or six closely related keywords. A knee replacement ad group holds searches like "knee replacement surgeon," "knee replacement surgery," "knee replacement doctor" and "partial knee replacement." ACL reconstruction gets its own group. So does rotator cuff repair.
Then every ad in the group says what the group is about. The knee replacement ads talk about knee replacement: the procedure, the surgeons who do it, the recovery, the insurance you take. When the search, the keyword and the ad all say the same thing, ad relevance takes care of itself. That's the cheapest improvement available in the whole account, and it costs nothing but an afternoon of reorganizing.
How finely to split depends on how differently people search. A useful order:
- Body part first. Knee, hip, shoulder, spine, hand and wrist, foot and ankle. Most practices can stop here for body parts they rarely treat.
- Then procedure. Within the knee, replacement, ACL reconstruction and meniscus repair are different patients with different questions, so each gets its own group.
- Then condition searches, kept apart from procedure searches. "Torn meniscus treatment" and "meniscus surgery" come from people at different stages. Chapter 6 covers which of these belong in your first campaign.
A group that holds searches for two different procedures is already too loose. If you can't write one headline that fits every keyword in the group, split it.
A Knee Ad Should Land on a Knee Page
The page is the other half of the match. Send a knee replacement ad to the home page and the patient has to go looking for the thing they asked about. Google's page grade notices.
The example sites show how uneven this is. Three of the twelve have a page for each condition or procedure, so a knee replacement ad has an obvious place to land. Eight have body-area pages, such as a knee page, that cover some procedures but not all of them. One has only three broad pages covering whole regions of the body. The practices with a page per procedure can match every ad group to a page. The others end up sending several different searches to the same general page.
A good knee replacement page, for ad purposes, answers the searcher's first four questions before they scroll:
- What is this? The procedure named in the headline, in the patient's words.
- Who does it? The surgeons who perform it, with their credentials.
- Will my insurance work? The plans you accept, or a line on how to check.
- How do I start? A phone number and a way to request an appointment, visible on a phone screen.
If your site doesn't have pages like that yet, build them for the procedures you want most before you spend on those ad groups. An ad group without its own page is only half built. One page can serve two very close groups, say total and partial knee replacement, if it genuinely covers both. It shouldn't serve the knee and the shoulder.
Seven of Twelve Display Paths Promised Pages That Weren't There
Search ads show a web address under the headline, and part of it, the display path, is text the advertiser writes, like /knee or /dr-smith. It's meant to preview where the click goes.
On seven of the twelve advertisers' websites, the display paths their ads used led to a "page not found" error or a redirect when we typed them in. That doesn't prove the ads send people to broken pages, because the display path doesn't have to match the real destination, and the real destination isn't public. It does show ads written without a real page behind the promise. A patient who sees /knee in an ad expects to land on something about knees, and the fix is simple: make the display path describe the page the ad actually sends people to, and make sure that page exists. Websites change, too. Pages get renamed in a redesign and old addresses quietly break, so check every ad's final URL and display path whenever the site changes, and at least once a quarter even when it doesn't.
Speed Is Part of the Page, Not a Separate Project
How the page behaves on a phone counts too, and that's where most orthopedic searches start. Google's free PageSpeed Insights tool scores any page from 0 to 100 on mobile performance. The twelve example home pages ranged from 32 to 98, with most in the 40s to 70s.
Run your own procedure pages through it. If they score poorly, the usual fixes are the unglamorous ones: smaller images, fewer scripts and chat widgets loading at once, and nothing that shifts the layout while the page loads. But keep speed in proportion. A fast page about the wrong procedure still loses to a slower page about the right one. Match first, then speed.
Next: Filling Each Ad Group With the Right Searches
You have the structure now: ad groups by body part and procedure, each with its own ads and its own page. Chapter 5 is about filling those groups with the searches worth paying for, using the websites of the practices you compete with to find them, and keeping out the searches that aren't.
For the budget math behind all of this, see Chapter 3. The guide home lists every chapter in order.
Our Google Ads management work for orthopedic groups starts with this rebuild: body parts into ad groups, ad groups into matching pages, before any budget goes up.
