AEO for therapists: the directory is winning the answer.

Ask an assistant to find you a therapist and watch where the answer comes from. It is almost never a practice website. It is a directory, or a national teletherapy service, or a health publisher — because those are the sources that have published the explanations, and a practice site that is a bio, a fee, and a contact form has published none. AEO for therapists is therefore a different problem from every other vertical on this site: the competitor is not the practice down the street. It is an aggregator, and the practice is not currently a candidate.

Why the aggregator wins by default

Directories solved a real problem and they solved it well: a structured, filterable list of clinicians with modality, insurance, and availability in fields. That structure is precisely what makes them so easy for an engine to answer from. A directory can say “here are four EMDR-trained clinicians in this city taking new clients” because it holds that as data.

What a directory cannot do is explain. When the question is what EMDR actually involves, or how out-of-network reimbursement works, or how to tell whether a couples therapist is a fit, the engine needs prose written by somebody who knows — and it currently takes that prose from national publishers and teletherapy marketing. That is the gap. It is not filled by being listed more places; it is filled by publishing the explanation on a site you own.

Modality is the filter that most practice sites fail

People searching for therapy in 2026 arrive with vocabulary. They ask for EMDR, IFS, DBT, somatic work, ACT — by name — and a practice site that says “I provide a warm, collaborative space” cannot be matched to any of it. Naming the modalities you actually practice, and describing what a session in each one is like in practical terms, requires no efficacy claim at all and is the single most matchable thing a clinician can publish. What those modalities treat, and how well, is a clinical question for you and your board; nothing on this page addresses it and nothing prepared for a practice should.

Access is where the search dies

Whether you are taking clients. What you charge. Whether you are in network, out of network, or offer a sliding scale, and what out-of-network reimbursement actually involves for someone who has never done it. These are the most common reasons a search stops, and they are the questions practice sites are quietest about. Publishing them plainly — including “not currently accepting, as of this month” — is more useful to the person searching than silence, and it is the kind of specific, checkable fact an engine will relay.

Fit is the reason people stop and restart

The questions in the third row are about being understood, and they are the ones people are most reluctant to ask a human. An assistant gets asked them at eleven at night by someone who has been putting this off for a year. Describing who you work with — specifically, not as a list of adjectives — is how a practice becomes an answer to those, and it is also how the right clients find you rather than the nearest ones.

The behavior underneath the search

What gets built, and in what order

• One page per modality you practice. What it involves, what a session is like, what training you hold in it. The most matchable content in the vertical, and no efficacy claim required.
• The access page. Availability with a date, fees, insurance position, what out-of-network reimbursement involves, sliding scale if you offer one.
• Who you work with, described specifically — populations, situations, life stages — rather than as a list of affirming adjectives.
• The first-session page. What happens, how long, what to expect, how switching therapists works. Quietly one of the most-asked and least-answered questions in the category.
• The entity pass. Practice name, clinician names, license type and state, office location and telehealth coverage, stated identically everywhere they appear, with organization and person schema matching the visible page. Google states no special markup is required to appear in AI Overviews or AI Mode — schema is for unambiguous parsing and classic rich results.

Everything arrives as a draft in an approve queue — one prepared move per day, nothing published until the clinician approves it. In a licensed practice that review is the point, not an obstacle.

How you know whether it worked

Scans re-ask the same client questions on a schedule, so the record is a comparison rather than a promise. Approved pages go to Bing within minutes via IndexNow, and Bing standing is the fastest indicator — independent testing found roughly 87% of ChatGPT's citations match Bing's top results (Seer Interactive, 2026), a measured correlation and not an OpenAI-documented mechanism. Citations run to a median of about 6.8 days after indexing, with roughly 42% of pages cited within 30 days, in practitioner testing across 2026 (Semrush / practitioner testing, 2026). Nobody honest promises a date, and 40–60% of the sources AI cites change every month (eMarketer, 2026).