Sixty-eight percent of Google searches in the US now end without a click.
That's SparkToro's number for the first four months of 2026, up from about 60% in 2024. A patient types "is Botox or filler better for forehead lines," reads what Google puts at the top of the page, and closes the tab. They got their answer. Nobody's site got a visit.
The reflex is to treat that as a loss. It's worth slowing down on, because the number underneath it points somewhere more useful.
The stat everyone quotes, and the one that actually matters
You'll see "AI Overviews appear on half of all searches" repeated a lot. Be careful with it. The real range depends entirely on which keywords a study tracked: BrightEdge put it around 48% in early 2026, Xponent21 around 60%, Conductor around 25%, and Semrush closer to 16%. Informational questions trigger them constantly. "Med spa near me" often doesn't.
So the honest version is: on the treatment questions your future patients ask before they're ready to book, an AI summary is very likely sitting on top of the results. On the searches where they're ready to call, it's less of a factor — that's still map pack territory.
Here's the number worth building around instead. Seer Interactive measured brands cited inside an AI Overview against brands that ranked on the same query but weren't cited. The cited ones earned roughly 120% more organic clicks per impression. Same query. Same page. The difference was being named.
So the goal was never to beat the summary. It's to be in it.
What makes a page quotable
An AI summary is assembled from passages it can lift cleanly. That's the whole mechanic, and it explains most of what follows.
It answers the question in the first two sentences
Not after a paragraph of throat-clearing about your commitment to patient care. If the page is about Botox for forehead lines, the first thing on it should tell you what Botox does to forehead lines.
It says who the treatment is for, and who it isn't
Candidacy language is the single most under-used thing on aesthetic websites, and it's exactly what these systems reach for. "May be appropriate for patients who…" is quotable. "Perfect for anyone who wants to look their best" is not — it says nothing a summary can use.
It sets expectations with real numbers
How long the appointment runs. When results show. How long they last. What downtime looks like. Ranges are fine and more honest than a single figure. Vagueness is what gets a page passed over.
It's attached to a credible person
Who performs it, what their credentials are, how long they've been doing it. A page that names a provider carries more weight than one written by nobody in particular.
The same paragraph, twice
This is easier to see than to explain. Here's copy of the kind that sits on a lot of medspa treatment pages:
Our state-of-the-art Botox treatments are designed to smooth away those unwanted forehead lines and help you look refreshed and confident. Our expert team will create a personalized plan built around your unique needs, so you can walk out feeling like the best version of yourself. Book your consultation today and let us help you rediscover your natural beauty.
There is nothing in there a summary can use. No treatment mechanism, no candidacy, no timeline, no provider. Every sentence is about how the practice feels about itself.
Same treatment, written to be quotable:
Botox relaxes the muscles that create horizontal forehead lines and the vertical lines between the brows. It's typically most effective on dynamic lines — the ones that appear when you raise your eyebrows or frown — rather than lines that stay visible at rest. Most patients see softening within 3 to 7 days, with full results around two weeks, lasting roughly 3 to 4 months. Appointments usually run 15 to 30 minutes with no downtime. It may not be appropriate for patients who are pregnant or nursing, or who have certain neuromuscular conditions, which is why every patient is evaluated by our injector before treatment.
The second version is longer, less flattering, and far more useful. It also answers four separate questions a patient might type, which means it can be pulled into four different summaries. And you'll notice it never promises a result.
Why this argues for fewer pages, not more
The instinct when search changes is to publish more. This particular change argues the opposite.
A summary quotes depth, not volume. Ten thin treatment pages that each say "we offer this, call us" give Google nothing to lift. Three pages that fully explain a treatment — mechanism, candidacy, timeline, aftercare, cost range, provider — give it a great deal.
That reallocation is uncomfortable because it means fewer things published and more time per thing. But content built to actually explain your services is doing double duty now: it's what converts a hesitant patient and it's what gets you cited. It's the same work.
One category worth pulling forward: pricing. "How much does Botox cost" is one of the highest-volume questions in aesthetics and one of the most avoided pages. We've made the case for cost pages before, and AI summaries make it stronger — a page with an honest range gets quoted, and a page that says "call for pricing" gets skipped.
The caution that's specific to our field
Everything above pushes toward specificity, and specificity in aesthetics has a line you shouldn't cross.
Be concrete about mechanism, timing, and candidacy. Stay careful about outcomes. "Results vary" isn't legal boilerplate you tack on at the end — it belongs in how the page is written. A page that says a treatment "may be appropriate for patients who…" and "most patients see…" is both more responsible and more quotable than one making promises. That's a rare case where the careful version and the effective version are the same version.
And no page should imply an evaluation isn't necessary. Your injector deciding someone isn't a candidate is a feature of your practice, and saying so out loud builds more trust than any headline about confidence.
One page, one hour
If you want to try this on your own before deciding it needs help, don't audit your whole site. Take one page.
Pick the treatment you get asked about most. Not your highest margin — your most-asked. That's where the questions already exist.
Write down the five questions patients actually ask you about it at the front desk. The real ones, including the awkward ones about pain and cost.
Open the page and find the answers. If a question isn't answered, that's a gap. If it's answered in a way that hedges without saying anything, that's also a gap.
Rewrite the opening two sentences to answer the main question directly. No preamble. This one change does more than anything else on the list.
Add an FAQ block at the bottom with those five questions as headings. Question as the heading, direct answer underneath, two to four sentences each. This is the format these systems lift most readily, and it's the fastest win available to you.
Then leave it alone for six to eight weeks. Google has to recrawl, and nothing you do will speed that up.
If that hour never appears
Rewriting one page is a manageable afternoon. Rewriting thirty, in a voice that's consistent and medically responsible, while running a practice, is a different thing entirely.
That's the work we do for aesthetic practices — AI search optimization and treatment content built the way described above, for medspas and plastic surgery practices alike, usually under a monthly retainer.
We've also built an AI agent for this exact problem. Point it at one of your pages and it does two things: it reads the schema already available for that page and identifies content that could be added to match it, and it compares your page against the practices you're competing with to find the content gaps holding back your AI visibility. It's a much faster way to find out what a page is missing than reading it out loud and hoping you notice.
If you'd like to see the upgraded version of one of your own pages, get in touch and we'll run it for you.
