Hyphen

Search is becoming a distributed decision environment

Patients still use Google, maps, reviews, clinic websites, referrals, communities, and video. AI-generated answers are becoming another research layer, and the advertising systems are reorganizing around it. This memo tracks where each surface is heading and what a fertility business should build now.

By HyphenPublished Last materially reviewed

The research is moving into the answer

A growing share of patient research now ends inside the answer, before any clinic page loads.

AI Overviews respond directly at the top of the results page, and assistants answer in conversation, so reading that once produced a website visit increasingly produces none.

In the first four months of 2026, roughly 68 percent of U.S. Google searches ended without a click to the open web.SparkToro (Similarweb clickstream data), 2026

When an AI summary appears, users click a traditional result in about 8 percent of visits, versus 15 percent without one.Pew Research Center, 2025

Demand did not shrink; the visit that used to carry the research disappeared. The visits that remain arrive later in the decision and carry more intent than the old top-of-search click ever did.

The ad is moving into the answer

Organic visibility inside AI answers is unstable. Paid placement is the part of the answer layer a clinic can decide to hold.

Sponsored placements render inside Google’s AI experiences and Microsoft’s Copilot answers, and the assistants are building commercial formats. As the earned side of search gets less predictable, the controllable side matters more. What qualifies a business to use it well has not changed: a legible site, honest conversion signals, and an offer clear enough to win a decision that is already half-made by the time the click happens.

Where each surface is heading

Every surface is converging on answer-shaped results. The preparation is the same for all of them.

Rendered from structured direction data. Each entry carries its own review date and sources. Methodology: Direction entries are compiled from platform documentation and announcements, published research on search behavior, direct observation inside managed accounts, and industry reporting. Each entry links its sources and carries its own review date. Entries are re-verified on a monthly cadence or when a platform announces a material change, whichever comes first.

Google search

The results page is becoming an answer surface

AI Overviews sit above the links and answer a growing share of queries directly, with AI Mode extending the same behavior into full conversation. Ads are following the answer: paid placements render within and alongside AI experiences, served through Search campaigns rather than a separate buy. The page a patient sees is increasingly an assembled response, not a ranked list.

For fertility businessesHealth queries receive more conservative AI treatment than retail, and the high-intent provider-selection searches that decide clinic revenue still resolve through search. The surface is changing shape faster than the decision behind it.

Reviewed · Google Search Central documentation · Pew Research Center

Google Ads

The account is becoming a relevance engine steered by the signals you feed it

Google's own direction is explicit in AI Max for Search: matching expands past your keyword list toward the intent behind queries, ad copy is assembled from your ads and landing pages, and the system chooses which page each search lands on. Every one of those automations is steered by two inputs the advertiser controls: the conversion signal returned to the account, and the clarity of the website it reads.

For fertility businessesThis is where measurement stops being reporting and becomes the steering input. An account fed form-fills optimizes toward inquiry volume; an account fed consultation and treatment-start signals optimizes toward patients. The same automation, pointed at different definitions of success.

Reviewed · Google Ads Help

Microsoft

The same direction on the second surface

Bing and Copilot are moving the same way Google is: answers first, sponsored placements inside them, automation reading the advertiser's site. Microsoft matters less for volume and more as confirmation that the whole category of search advertising is converging on answer-shaped results.

For fertility businessesA disciplined Microsoft account is inexpensive coverage of the same intent, and it inherits the same requirement: honest conversion signals and legible pages.

Reviewed · Microsoft Advertising

ChatGPT and AI assistants

The research layer patients reach before any clinic page

Patients ask assistants the questions they used to research across many tabs: how treatment works, what it costs, how to compare providers. Sponsored formats are emerging here too. OpenAI has moved from testing ads toward self-serve formats, and every major assistant is experimenting with commercial placements. Assistant answers will carry paid presence, and the answers themselves are assembled from content the model can parse and trust.

For fertility businessesThe durable preparation is the same work that serves patients: clear, sourced, dated pages on the questions they actually ask, and consistent entity information everywhere the clinic appears. Vendors selling guaranteed assistant visibility are selling something no one controls.

Reviewed · OpenAI · Hyphen

Reddit and communities

Candid research that now feeds the answer layer

Fertility research on Reddit is unusually honest: costs, experiences, clinic names. Community content increasingly surfaces inside search results and AI answers, which makes these threads part of the research layer even for patients who never open Reddit. Paid placement exists and works selectively; the larger shift is that what communities say about a clinic travels further than it used to.

For fertility businessesWorth attention when the communities a clinic serves are active and the clinic has something genuinely useful to contribute. Sensitive-category context demands more care here than anywhere else.

Reviewed · Reddit for Business · Hyphen

YouTube and video

Treatment education is becoming a video behavior

Patients research procedures, clinics, and experiences in video, and video results surface inside search and AI answers. As a paid surface it rewards clinics with real video assets and patience for a longer path to consultation; as a research surface it shapes decisions whether or not a clinic buys it.

For fertility businessesJustified where genuine video exists and the economics tolerate demand creation. The same measurement standard applies: no downstream evaluation, no budget on faith.

Reviewed · Google Advertising Policies Help · Hyphen

What clinics should do now

Citations are influence, not control. The durable work below compounds across every surface, current and future.

Which sources get cited inside AI answers is unstable, only loosely tied to rankings, and subject to interface changes no one outside the platform can predict. The fundamentals are worth doing for their own sake, and any promise of guaranteed answer placement is the tell that it is one.

  • Clear, useful clinic pages that answer real questions
  • Accurate entity and provider information everywhere it appears
  • A strong local presence: profiles, reviews, and consistent details
  • Search-to-page relevance maintained per intent
  • Source-backed educational content with visible dates and authorship
  • Structured content and metadata that machines can parse
  • Measurable consultation paths from every entry point
  • Conversion signals that reflect patients, not inquiries

Change log

Initial publication.

Reframed the page around where each channel is heading, replacing the inventory-status ledger. Added the paid-search automation thesis.

Compressed the memo for skimming: takeaway lines added, contents collapsed to four sections, and the practice-limits list consolidated onto the paid acquisition page.

Start with the path from click to patient outcome

The diagnostic shows what the clinic is currently counting, where attribution breaks, what may be transmitted, and which changes should happen first.