People describe the problem before they name a therapist.
before they name a therapist
Someone types four paragraphs about what is happening at home, then asks who could help. We run ChatGPT Ads and answer visibility for therapy practices so your practice is described accurately in that conversation, and the enquiry reaches a real intake process. Commercial-intent conversations inside ChatGPT rose from 13.9% to 19.2% in a year.
Tell us a little about your brand and we'll be in touch within 24 hours to lock in a time.
FOUR WORKSTREAMS
Four workstreams behind assistant advertising for a therapy practice.
a therapy practice
The questions people actually type, advertising that stays inside health rules, a published record an assistant can read, and measurement that never touches clinical detail.
Question and demand mapping
Compliant placements
Published record and answer visibility
Private, honest measurement
What someone types at eleven at night.
These are not keywords. People write paragraphs: what a panic attack feels like, whether their insurance covers weekly sessions, whether EMDR is different from talk therapy, whether they can see someone before a work trip. An analysis of 36,153 answers found even purely informational questions drew an advert 69.5% of the time, so those conversations are already commercial ground.
We map that question set by modality, by payer situation and by who is asking — the person themselves, a partner, a parent, a referring GP — then decide where a placement is appropriate and where clear published content on your own site does the work instead.
- Question sets built per modality and presenting issue
- Client, family and referrer conversations separated
- Fees, sliding scale and insurance questions mapped honestly
- Crisis-language prompts excluded from advertising
19.2%
of ChatGPT conversations now carry commercial intent
Health review settled before any spend.
Mental health advertising sits under stricter review than most categories, and some modalities carry extra requirements: psychiatry and medication management campaigns run $70-$140 per click and need certification before they serve at all. We settle the review position, the claim language and your clinical sign-off route before a single impression is bought.
Then we buy. Placements are set against the questions where an advert genuinely helps somebody reach care, budgets follow the hours a human answers your phone, and every line of copy names the modality, the location, the fee position and the waiting time rather than promising anyone how they will feel.
- Health review position confirmed before launch
- Claim language reviewed by your clinical lead
- No outcome or symptom-improvement claims, anywhere
- Budgets aligned to staffed intake hours
3,770
active advertisers on ChatGPT in one measured week
46%
week-over-week growth after the spend floor dropped
Assistants describe you from what you publish.
When somebody asks for a trauma therapist in your city, the model assembles an answer from your practice pages, your directory profiles, your licence records and your reviews. If your modalities, licences, fees, telehealth states and waiting list are not written plainly, the description is vague or simply wrong — and you never see it happen.
We fix that record with our AI search visibility and SEO teams: accurate clinician pages, structured data, consistent directory details, and a monthly check on how the major assistants describe your practice.
- Modalities, licences and states written plainly
- Fees and insurance position published clearly
- Directory and listing details made consistent
- Assistant descriptions monitored for accuracy
5.8%
average search ad click-through rate in mental health
4.2%
average conversion rate on those clicks
Measured to a booked first session.
Paid demand here is not cheap once you count the people who never book. Mental health click-to-form-fill cost per lead runs $110-$155, and specialty modalities such as EMDR, OCD and trauma work sit at $75-$175 per lead. What matters is the booked, attended first session.
We measure to that, using call tracking and form attribution configured so no presenting issue, diagnosis or clinical detail ever reaches an advertising platform. Our analytics and privacy teams set that up together and you get a written monthly read.
- Call and form tracking configured for health privacy
- No clinical detail passed to advertising platforms
- Cost per enquiry and per attended session reported
- A written monthly read, not a dashboard link
$155
upper benchmark cost per mental health lead on search
$3.85
average mental health search click in 2026
upper benchmark cost per mental health lead on search in 2026
average mental health search ad click-through rate
Written read on enquiries and booked sessions
Ad accounts, content and tracking in your name
We made the difference for those brands
Retail & commerce
Consumer tech and platforms
B2B software, fintech, insurance
Consumer tech and platforms
Retail & commerce
Food & beverage
Retail & commerce
Healthcare & regulated services
Home essentials, appliances, kitchen & pet
Healthcare & regulated services
Creative, content, arts & culture
Healthcare & regulated services
Consumer tech and platforms
Beauty, personal care & wellness
Apparel and lifestyle
Creative, content, arts & culture
Apparel and lifestyle
Retail & commerce
Consumer tech and platforms
Beauty, personal care & wellness
B2B software, fintech, insurance
B2B software, fintech, insurance
Food & beverage
Home essentials, appliances, kitchen & pet
Beauty, personal care & wellness
Home essentials, appliances, kitchen & pet
Healthcare & regulated services
Beauty, personal care & wellness
Healthcare & regulated services
B2B software, fintech, insurance
01 — The challenge
A new front door to a practice that only has so many hours.
Therapy practices do not need unlimited enquiries. They need the right ones, reaching a clinician with capacity, at a fee that works. Search already charges for the rest: mental health leads cost $110-$155 apiece even before the no-shows. Meanwhile the first conversation is increasingly with an assistant, and 72% of US ChatGPT answers already carry an advert.
“A client told us the chatbot had recommended us. It also told her we took her insurance. We do not.”
That is the real risk: not silence, but a confident description nobody at the practice has read. A fifth of these conversations now carry commercial intent.
02 — Our approach
Accuracy first, review second, spend last.
We start by recording how the major assistants describe your practice today: which clinicians they name, which modalities they attribute to you, what they claim about fees, insurance and availability. That transcript is usually the most useful document in the engagement, because it shows precisely where your published information is thin, dated or contradictory.
Then we correct the record. Clinician credentials, licensure states, modalities, telehealth availability, fee ranges, sliding scale and the actual intake steps get written plainly and marked up properly, so an assistant can describe you without guessing. That work lifts ordinary search too, and it costs nothing per click.
Only after the health review position and clinical sign-off are settled do we buy placements, against the questions where an advert genuinely helps someone reach care rather than intercepting a person in crisis. Tracking is configured with your privacy obligations first and intake reporting second, and you receive a written monthly read. Accounts, content and data stay in your name throughout.
03 — What we did
Record, correct, review, then buy.
How assistants describe your practice today, the published record corrected, health review settled, and placements measured to a booked first session.
Weeks 1-2 / Baseline
How assistants describe you today
Answers recorded for each clinician, modality and city, with every inaccuracy about fees, insurance and availability documented before anything changes.
Weeks 2-6 / Evidence
The practice record, written plainly
Credentials, licensure states, modalities, fees and intake steps published clearly and marked up so models read them correctly.
Month 2 / Review
Health review and clinical sign-off
Ad review position confirmed, claim language approved by your clinical lead, privacy-safe tracking configured before launch.
Ongoing / Prove
Cost per attended session, reviewed monthly
Enquiries and booked sessions attributed by source and read against your existing search, directory and referral channels.
WHAT YOU GET
Deliverables you keep, whatever happens next.
you keep
The ad accounts, the practice content, the tracking configuration and the documentation are in your name from the first week.
Assistant description audit
What ChatGPT and other assistants say today about your practice, clinician by clinician, with every inaccuracy listed.
Client question map
The questions clients, partners and referrers ask, grouped by modality, payer position and urgency.
Practice content rebuild
Clinician credentials, licensure states, fees and intake steps written plainly and marked up properly.
Ad review pack
Health review position, approved claim language and a documented clinical sign-off route for every advert.
Privacy-safe tracking
Call and form attribution configured so no presenting issue or clinical detail reaches an ad platform.
Monthly reporting pack
Cost per enquiry and per attended first session, with a written note on what moved and what happens next.
HOW WE WORK
Operating standards, not promises.
Operating standards
Solo and small practices
One or two clinicians with limited hours, where the wrong enquiry costs more than an empty slot.
ExploreGroup practices
Multi-clinician teams matching enquiries to the right modality, licence state and waiting list.
ExploreReferral partners
Physicians, schools and employers who need your modalities, fees and intake route stated precisely.
Explore
Built on trust. Proven by results.
We partner with SMBs and Fortune 500 companies to deliver more than reach — we bring clarity, execution, and measurable outcomes. Every successful partnership starts with a strong culture fit and a shared drive to grow.
CASE STUDIES
Case studies
Website Portfolio
Video Ads
Static Ads
EN
EN
EN
FR
FR
FR
FR
FR
FR
FR
FR
EN
FAQ
Questions therapy practices ask us
Can a therapy practice advertise inside ChatGPT at all?
Yes, within health advertising rules that are reviewed case by case, and we handle that review before anything runs. Some adjacent services carry extra requirements — psychiatry and medication management campaigns need certification and run $70-$140 per click. Where a placement is not appropriate, we say so plainly and put the budget into visibility work that keeps paying either way.
What if ChatGPT describes our practice inaccurately?
That is the most common finding, and it is fixable. Assistants assemble descriptions from what you publish, from directory profiles and from licence records, so a thin clinician page produces a vague or invented answer. We record the current descriptions, list every inaccuracy about modalities, fees, insurance and availability, then correct the underlying record so the description improves at source.
How do you protect client privacy in tracking?
Privacy configuration comes before measurement. Call tracking and form attribution are set up so no presenting issue, diagnosis or sensitive parameter is passed to an advertising platform, and our privacy team documents the setup for your records. Reporting works at the level of enquiry sources and booked sessions, which is enough to run a budget responsibly.
Will you make claims about how people will feel?
Never. We do not write recovery, symptom-improvement or guarantee language in any format, because it is unverifiable, it breaks platform policy and it damages the trust someone arrives with. Copy describes modalities, credentials, licensure states, telehealth availability, fees, sliding scale and how long the wait is — concrete facts your clinical lead can approve line by line.
What does this cost compared with Google Ads?
Search is the reference point: mental health clicks average $3.85 with a 4.2% conversion rate, which lands cost per lead at $110-$155. Assistant placements are far less contested today, which is exactly why testing them now is sensible rather than brave. We run both and let cost per attended session decide the split each quarter.
We only have a few open slots. Is this worth it?
Often more worth it than broad advertising, because the work is about precision rather than volume. A practice with six open hours a week wants those hours filled by clients whose modality, fee position and location actually match. Mapping the questions and correcting your published record does that filtering before an enquiry ever arrives, and you can pause the paid half at any point.
How do you handle someone in crisis?
Crisis language is excluded from advertising entirely. We build an explicit prompt and keyword exclusion set covering self-harm, suicidality and emergency phrasing, so no advert of ours competes for those conversations. Your published content should still carry clear crisis-line signposting, and we make sure it does as part of the content work, because it is the right thing and assistants read it too.
How long before this shows results?
The visibility work usually shows within a quarter, because assistants re-read your published record as it improves. Placements can produce enquiries in weeks. We report the two separately so you always know which part is paying, and in the first month we would rather show a small number of well-attributed booked sessions than a large click count nobody can trace.
Do we need to change our website for this?
Usually a section of it, not all of it. The pages that matter are the clinician profiles, the modality pages, the fees and insurance page and the intake instructions, and those normally need rewriting for clarity rather than redesigning. Where a rebuild genuinely helps, our web design team scopes it separately so you can decide on its own merits.
How does this relate to our SEO and directory listings?
It shares a foundation. Accurate clinician content, consistent directory profiles and structured markup are what both an assistant and a search engine read, so practices already running SEO start well ahead. We run one content plan across both rather than duplicating work, and report assistant visibility alongside ordinary rankings and directory performance.
Are other practices already doing this?
Very few, which is the opportunity while it lasts. Active advertisers across every category rose from 2,585 to 3,770 in a single week worldwide — a tiny field beside the number of therapy practices in one US state. Behavioural health is barely represented in assistant advertising today, and early accuracy work compounds.
Who will be working on our account?
A named senior strategist with a healthcare compliance specialist and a content lead alongside, and the same people stay with your practice. You will know their names, you can email them directly, and the person who wrote your plan is the person reviewing the numbers with your intake coordinator each month.
.webp)
.webp)
