Everything in this guide so far waits for a patient to type something. Search ads answer a search. Performance Max leans on the same intent. Demand Gen is the first campaign type that goes out and finds people before they've typed anything at all, while they're watching YouTube, scrolling Discover or reading their email.
That's a different job, and in healthcare it comes with a twist. For most of a practice's service lines, Google's health rules shrink Demand Gen down to one place: YouTube. So a healthcare Demand Gen campaign is really a video campaign, and the video decides whether it works. Plan the video first. The campaign settings take an afternoon.
Demand Gen Reaches Patients Before They Search
Google built Demand Gen to create interest, not to answer it. Its ads can run across YouTube, including Shorts, the Discover feed in the Google app, Gmail and the Display Network. Instead of keywords, it targets people: audiences Google has built around interests, life events and what people are in the market for.
That's why it comes late in the rollout this guide teaches. The Search account from Chapters 5 through 9 catches the patients who are ready to book. Performance Max, from Chapter 10, adds reach once Search has real bookings. Demand Gen reaches further up, toward people who aren't looking yet: the allergy patient a few weeks before the spring rush of searches, the parent who hasn't picked a pediatrician, the adult who's been meaning to see someone about their knee.
Those people won't book from one ad. The point is that when they do search, weeks later, your practice isn't a stranger. That makes Demand Gen the last campaign a practice should fund, not the first. It still reports to the same goal as everything else. The bookings Chapter 1 set up stay the conversion Demand Gen optimizes toward, so it's judged on patients, not on views.
For a Sensitive Service Line, Demand Gen Is a YouTube Campaign
Chapter 10 split a practice's service pages into sensitive and everything else. That sort matters even more here.
Google's help center says ads in sensitive categories are generally restricted to YouTube inventory and aren't eligible to serve on Gmail or the Display Network. And the only people they may aim at are the ones Google itself has grouped, such as its in-market and affinity audiences. Nothing the practice assembles on its own qualifies. Discover is possible in some cases, but Google's own help pages describe the conditions differently, so don't count on it. Pediatric practices face one more wall. Google's health policy keeps anyone under 18 out of personalized ads entirely, Google-built audiences included. A pediatric campaign reaches parents, so the video should speak to parents.
So a colonoscopy campaign, a psychiatry campaign or a fertility campaign in Demand Gen runs mainly on YouTube. A non-sensitive line, such as urgent care or a primary care page inviting new patients, keeps more of the inventory. The same practice can end up with two very different Demand Gen campaigns: one that shows up in inboxes and feeds, and one that lives almost entirely in video.
That's the real reason the video comes first. For most sensitive lines, there's no image ad running quietly in Gmail to carry the campaign while the video gets made. If there's no good video, there's no campaign. A practice that isn't ready to film should wait.
The Video Is the Ad, So It Gets Made First
A practice doesn't need a production company. It needs a real doctor, the real office and about thirty seconds of plain talk. The goal is to make a stranger feel like they've already been inside.
What works in healthcare is the same thing patients praised in the reviews from Chapter 8: someone who listens, a front desk that answers, a visit that isn't rushed. Show that. A doctor introducing themselves and saying what a first visit is like. The front desk, smiling. The room where the patient will actually sit. How to book. Nothing more. Fifteen to thirty seconds is plenty for the main version, with a shorter cut for Shorts. Make one video per service line rather than one for the whole practice, for the same reason Chapter 8 gave each ad group its own voice: the orthopedic patient and the allergy patient want to hear about different things. And put the same video on that service's landing page from Chapter 4. A patient who clicks through should see the doctor they just watched.
A few rules keep it honest and usable:
- Film the real people, with their permission. Stock footage of a stranger in a white coat looks like every other practice's ad.
- Say who you are in the first few seconds. Many viewers can skip, so the practice name and the service belong at the start, not the end.
- Add captions. A lot of video is watched with the sound off, and captions carry the message either way.
- Shoot for both shapes. Film so the video works wide for regular YouTube and tall for Shorts on a phone.
- Keep patients out of it, unless a patient has agreed in writing to appear. Never show anything that could identify someone's condition.
- Watch the tone. As Chapter 10 noted, Google treats ads that suggest bad things will happen if you don't act as a sensitive category. Invite; don't frighten.
- Don't promise outcomes. "Pain-free in a week" is a claim the practice would have to stand behind for every single patient who watches it.
And don't let Google make the video for you. If a campaign has no video, Google can assemble one from your images. It's rarely something a practice would choose to put its name on.
LASIK Searches Jump Tenfold From December to March. Urgent Care Barely Moves.
Demand Gen works best when it runs ahead of demand, reaching people just before they start searching. The timing is different for every service line, and a practice can see it in Keyword Planner's month-by-month numbers.
Here's what twelve months of searches looked like in our example market, San Diego:
- "lasik eye surgery san diego" ran 260 to 480 searches a month from October through December, then 2,400 to 2,900 a month from February through May.
- "allergist san diego" hit 1,300 in March, against 480 in the winter and late summer months.
- "primary care provider" ran 1,900 a month from January through April, against 1,300 in the summer.
- "fertility clinic" peaked at 1,000 in July, against 260 in December.
- "urgent care near me," "therapist near me" and "dermatologist near me" barely moved all year.
The data shows when patients search. It doesn't show why, and your market's calendar may differ, so pull your own numbers before you plan around them. The practical use is simple: a line with a clear season gets its Demand Gen campaign running a few weeks before the rise, so the video has done its work by the time people start searching. A flat line like urgent care has no season to get ahead of, and Demand Gen has a harder case to make there. In our example market, that means the LASIK video is finished in December and the campaign is running in January, as searches start to climb toward the February peak. A fertility campaign would be live by June, ahead of the July jump.
The same calendar applies to Search. A seasonal line needs more Search budget during its peak months too, and Chapter 3's budget math works month by month, not just once a year.
Demand Gen's Budget Covers One Service Line at a Time, Not Five
Google's budget guidance for Demand Gen is blunt: a budget of at least 10 times your target cost per conversion. For a line where a booking costs $150, that's $1,500 before the campaign has enough room to learn.
Most practices can't fund that for every service line, and they shouldn't try. Pick one line: the one with a clear season coming up, a new service to announce or a new location that nobody knows about yet. Run it for the stretch that matters, then decide from the results whether it earns a second run. If the budget can't reach that level, don't run a starved Demand Gen campaign that never gets enough room to learn. Put the money into that line's Search campaign for the season instead, where every dollar still buys a patient who's already looking.
Judge it against the rest of the account, not against itself. Demand Gen will report its own conversions, some of them from people who watched a video and later booked some other way. Chapter 7 warned about numbers that look better than they are. The more honest test is whether the Search account's bookings for that service line rise while the video runs, and whether searches for the practice's name go up with them. Write down both numbers for the month before the campaign starts, so there's a real baseline to compare against, and change as little else as possible while the video runs. If Search bookings and brand searches don't move after six to eight weeks, the video didn't reach the right people or didn't give them a reason to remember you, and that's worth knowing before the next season.
Our specialty guides cover Demand Gen in their own fields. The dermatology guide covers using YouTube to reach people before they think about a skin check. The plastic surgery guide covers Demand Gen and YouTube ads for a plastic surgery practice. The optometry guide covers reaching parents before back-to-school eye exams.
Next: Pay per Lead, With Far Less Control Over Who Calls
There's one more Google ad product built for local businesses, and it works on completely different rules. Chapter 12 covers Local Services Ads: which healthcare specialties can use them, and why the method this guide teaches treats them as a late, optional add.
Catch up on Chapter 10 on Performance Max without patient lists, or see every chapter on the guide home for Google Ads for Healthcare. When a practice hires us for Google Ads management, the video gets filmed in the practice's own rooms before a single Demand Gen dollar is spent.
