top of page
Alpine Analytix logo

Google Is Expanding AI Overviews Into AI Mode: Here's What Changes for Healthcare Marketers Across AEO, SEO, and Google Ads

Writer: Alpine Analytix
Alpine Analytix
Sep 2
7 min read
Ash Ledford, Co-Founder of Alpine Analytix




Illustration of a Google search results page where an expanded AI Mode answer with an Ask anything box pushes organic listings below the fold


On August 28, 2026, Search Engine Roundtable reported that Google confirmed AI Overviews now dynamically expand into a fuller AI Mode response for some queries, loading the "Ask anything" follow-up box by default and removing the "Show more" click. A Google spokesperson said the expansion happens "for topics where our systems determine it's most useful for people." Google's Robby Stein added that if a searcher has already begun scrolling, the expansion stops so the reader keeps their place.


For healthcare marketers, AI Mode changes the practical picture in four ways:


  • Organic listings get pushed materially further down the page on exactly the exploratory, multi-step queries that dominate healthcare search.


  • Visibility increasingly means being cited inside the answer rather than ranking above it.


  • Paid search takes on more weight at the point where a patient moves from researching a condition toward choosing a provider.


  • The campaign types that reach AI surfaces are automated ones, and they require offline conversion data, conversion volume, and active guardrails before they produce defensible results.



What AI Mode means for healthcare marketers


Previously, an AI Overview loaded as a snippet with a "Show more" button. A searcher had to click before seeing the full response.


Now, for a subset of queries, Google loads the complete AI answer immediately with the AI Mode prompt box already open. The ten organic results sit well below all of it.


Google has framed this as a dynamic experience applied selectively. Two details from that framing matter for planning:


  1. Expansion is triggered by topic complexity, described by Google as topics where its systems determine expansion to be most useful.


  2. Expansion is suppressed once a user scrolls, which tells you Google is measuring reading behavior and tuning against it.


Side-by-side comparison of the previous collapsed AI Overview layout and the new expanded AI Mode layout, showing organic results pushed below the fold

Healthcare queries sit squarely in the first category. Symptom questions, treatment comparisons, cost and coverage questions, and "what kind of specialist treats this" questions are all multi-step by nature. They are precisely the queries Google is most likely to expand.



What this means for AEO (answer engine optimization)


The unit of visibility becomes the citation. A patient reading a fully expanded answer with an open prompt box has little reason to click anything. Ranking in position one carries less weight when position one appears below the fold.


Optimize for the chain of questions, not the single query. Someone researching GLP-1 side effects will ask next about dosing, then cost, then insurance coverage, then what happens after stopping treatment. Google's AI surfaces explore related sub-questions to assemble an answer, so content organized around one keyword per page will lose to content organized around a full clinical decision path.


Healthcare organizations hold an advantage that is difficult to replicate. Named clinicians, stated credentials, medical review dates, first-party care protocols, and outcomes data belong to the organization delivering the care. Generic condition content cannot manufacture any of it. Health systems and multi-location groups that surface clinical authorship on every substantive page are giving Google's systems a reason to cite them.


Watch this space rather than buying promises about it. AI visibility for healthcare remains an evolving and restricted environment. Any partner guaranteeing placement inside AI answers is selling something Google has not made available.



What this means for SEO


Build the 2027 organic forecast on declining top-of-funnel traffic. Any model that extends last year's trendline forward is starting from a broken assumption. Informational content will absorb the largest share of the loss.


Value concentrates in three places. Appointment-intent and transactional queries, where the searcher must reach your site in order to act. Local and entity strength, since "who treats this near me" still resolves to a provider, a location, and a phone number. And branded query volume, which tends to rise as AI exposure builds recognition.


Branded query volume creates a measurement problem worth naming. A patient sees your organization cited inside an AI answer, never clicks, and returns days later by searching your name directly. That visit posts in analytics as branded organic or direct traffic. The investment that created the exposure receives none of the credit, and the reporting quietly understates it.

Google's Search Console AI performance reporting went live globally on September 1, 2026, which gives healthcare marketers a real data source for AI surface visibility instead of inference.



What this means for Google Ads


Google is running a small United States test of ads inside AI Mode for the healthcare vertical, on English-language queries. Google Ads Liaison Ginny Marvin confirmed the test on LinkedIn in late May 2026, around Google Marketing Live.


Marvin identified the eligible campaign types: Performance Max, AI Max with search terms matching, Shopping, and broad match. Those same campaign types were already eligible to serve in AI Overviews.


The pinned assets and disclaimer exclusion most coverage missed


Marvin noted one constraint that received far less attention than the headline. The first iteration of the test excludes creatives using pinned assets or text disclaimers.


That single line determines which healthcare advertisers can participate at all.


Regulated healthcare creative frequently depends on pinned headlines to keep required language in a fixed position, and on text disclaimers to carry mandated qualifications. Google made disclaimer fields available to regulated industries in Google Ads in April 2026, and then excluded exactly those formats from the first phase of AI Mode testing.


The implication is that Google's compliance layer for AI Mode ad formats is still under construction. Healthcare organizations whose compliance obligations live inside the ad copy itself are watching this phase rather than running in it. Those whose required language can live on the landing page instead have a narrower path into the test today.


Advertisers who wait for full format parity before testing will enter the channel with no performance baseline once it opens.


The trade nobody puts in the pitch deck


Automated campaign types are the entry ticket to AI placements. They also give Google broad discretion over where budget lands, including inventory and placements no marketing leader would have chosen deliberately.


Expanding from controlled search campaigns into AI Max, Performance Max, or broad match without preparation produces expensive noise. Three things belong in place first.


1. Offline conversion tracking, built HIPAA-safely. 

Booked appointments, consultations, and revenue events live in the EMR or practice management system, disconnected from the click that produced them. Until those outcomes flow back into the ad platforms, the algorithms optimize toward form fills, and form fills are what you will receive in abundance.


2. Sufficient conversion volume. 

These systems learn from data density. A campaign feeding thin or low-quality signals produces expensive guessing dressed up as machine learning. Our Required Budget Calculator shows whether a monthly budget generates enough conversion volume for the algorithms to learn from, or whether it is buying noise.


3. Active guardrails with a weekly review cadence. 

Placement and network exclusions, negative keyword governance, brand term controls, asset group discipline, and service-line economics somebody examines regularly. Automation without oversight will find the cheapest conversion available, and cheap rarely means valuable in patient acquisition.

Organizations that skip these steps generally discover the damage two quarters later, visible in cost per acquired patient rather than in cost per lead.



Frequently asked questions


Did Google remove the "Show more" button from AI Overviews entirely? 

Not entirely. Google confirmed the expanded experience applies to some queries, described as topics where its systems determine expansion to be most useful. The expansion also stops if a searcher has already begun scrolling.


Are healthcare ads running inside AI Mode right now? 

A small test is running in the United States on English-language queries, confirmed by Google Ads Liaison Ginny Marvin in late May 2026. Eligible campaign types are Performance Max, AI Max with search terms matching, Shopping, and broad match. Creatives using pinned assets or text disclaimers were excluded from the first iteration.


Should healthcare organizations move budget into Performance Max or AI Max now? 

Only when offline conversion tracking, adequate conversion volume, and enforced guardrails already exist. Those campaign types hand significant control to the platform, and control without measurement produces results nobody can defend in a board meeting.


How does an organization know whether its measurement can support automated campaigns? 

Start by auditing what exists today: how conversions are captured, whether patient outcomes reach the ad platforms, where HIPAA exposure sits in the current tracking setup, and where the connection between spend and revenue breaks. No two setups are the same, which is why an audit precedes any recommendation. Our healthcare measurement maturity guide includes a short self-assessment that maps where an organization sits today, with no email required.



Where Alpine Analytix fits


Alpine Analytix is a healthcare marketing agency built around a simple standard: every dollar you spend should be traceable to a result.


Our media team has spent years inside this regulatory landscape and knows what platforms will and will not approve. That work only compounds when the measurement underneath it holds up, so we audit your tracking and attribution, map the infrastructure your goals require, specify the right tools for your size, and get you tracking in a HIPAA-safe way. Then we optimize against real outcomes instead of clicks.


If your organization is weighing a move into automated campaign types while Google reshapes the search results page, the measurement question comes first. Talk with us about auditing what you have today.



About the author


Ash Ledford is Co-Founder of Alpine Analytix, where she leads client engagements for healthcare organizations and oversees day-to-day operations.


Before Alpine, she ran a health and wellness business, which shaped how she thinks about the gap between marketing activity and a booked appointment. She works directly with healthcare operators on what their measurement can support today, what it will take to reach the next step, and how to make a marketing investment defensible to the people who approve it.



Sources


Comments

bottom of page