Being the first telehealth brand ChatGPT names is probably worth little more than being named at all. The best available evidence, a June 2026 preprint from Scrunch AI, found AI recommendations lift brand searches and site visits about equally whether a brand is named first, second, or later. For a DTC brand, the target is share of answers.
AEO, GEO, AI visibility, and AI SEO are four labels for one discipline: getting a brand named and cited inside ChatGPT, Perplexity, Gemini, Copilot, and Google AI Overviews. This post uses AEO. The [plastic surgery version of this question](/blog/value-of-first-surgeon-chatgpt-names-plastic-surgery) prices a surgical case in one metro. A telehealth brand differs in three ways: the prompt set is national, the value is counted in CAC and patient LTV, and most of the measured effect arrives through a later Google search rather than a click inside the answer.
- A June 2026 preprint by Scrunch AI researchers, joining a clickstream panel to the same users' ChatGPT, Claude, and Gemini chats, found the off-platform effect of an AI brand recommendation is flat across mention order. Being named at all mattered; position in the list did not show a reliable gradient.
- In the same Scrunch AI study, an AI recommendation to a user with no recent engagement raised same-name Google searches by 4.3 percentage points and visits to the brand's own site by 2.4 points over the next 7 days.
- In a modeled national category with 20,000 relevant AI answers a month, each 10 points of mention share is worth about 3.8 paying telehealth patients a month, or $2,300 to $7,700 of lifetime value at an LTV of $600 to $2,000. These are modeled figures, not measured results.
- Because the measured lift runs mostly through Google search, a telehealth brand captures AI-answer value only if it owns its own brand-name results page. ClinicAds plans paid acquisition at $90 to $250 CAC, an agency average and not a guarantee.
What does being named first by ChatGPT mean for telehealth?
First-named means a telehealth brand appears first in the largest share of answers to a fixed national prompt set, such as best online TRT clinic or cheapest GLP-1 telehealth program. ChatGPT holds no fixed ranking, so the same prompt can return different brands in a different order on the next run, and a single answer proves nothing.
ClinicAds measures two numbers for a DTC brand: first-named share, the percentage of answers that list the brand first, and mention share, the percentage that name it anywhere. A working set is 24 prompts across the brand's conditions, price questions, and brand-versus-brand comparisons, each run 5 times, for 120 answers per engine.
Does being named first beat being named at all?
Being named first does not appear to beat being named at all, based on the one published study that tested it. Michael Iannelli and Alan Ai of Scrunch AI joined an opt-in clickstream panel to the same panelists' AI conversations and report that the off-platform effect of a recommendation is flat across mention order, with no reliable mention-position gradient.
Two caveats apply. The study is a non-peer-reviewed preprint from a company that sells AI visibility measurement, and it does not publish its panel size or name a health category. No telehealth-specific study of answer position was found while writing this post. The working conclusion is that mention share is the outcome to grow, and first-named share is a useful secondary reading of how strongly an engine associates the brand with a condition.
Where does the value of an AI mention show up?
The value of an AI mention shows up mostly in Google, not inside the AI answer. In the Scrunch AI study, a recommendation to a user with no recent engagement with the brand raised the following behaviors over the next 7 days, measured against matched earlier windows for the same users:
- 1. Same-name Google searches: up 4.3 percentage points
- 2. Visits to the brand's own website: up 2.4 percentage points
- 3. Visits to brand-specific retailer pages: up 1.0 percentage point
- 4. Incidental name-drops, where the brand was mentioned but not recommended, moved the same behaviors by roughly a third to a half as much
What is a share of AI answers worth in CAC and LTV?
The model assumes 20,000 relevant AI answers a month across a national telehealth category, that 2.4 percent of answers naming the brand lead to a site visit, borrowing the Scrunch AI lift as an illustrative rate, and that 8 percent of those visitors become paying patients. Every rate is an assumption and should be replaced with the brand's own intake and conversion data.
Two values are shown. Avoided CAC is what the same patients would have cost through paid media at ClinicAds' $90 to $250 benchmark. Lifetime value applies the $600 to $2,000 patient LTV band, which is the larger number and the one that justifies the work when retention holds. A brand spending $30,000 a month at $150 CAC acquires 200 patients; adding the 7.7 patients from 20 percent share lowers blended media CAC to about $144 before the AEO program's own cost, which the [AEO versus paid CAC comparison](/blog/is-aeo-cheaper-than-paid-ads-telehealth-cac) works through.
| Mention share | Site visits a month | Paying patients a month | Avoided paid CAC | Lifetime value of the month's patients |
|---|---|---|---|---|
| 10 percent | 48 | 3.8 | $346 to $960 | $2,304 to $7,680 |
| 20 percent | 96 | 7.7 | $691 to $1,920 | $4,608 to $15,360 |
| 30 percent | 144 | 11.5 | $1,037 to $2,880 | $6,912 to $23,040 |
How does a telehealth brand measure and raise its share?
A telehealth brand raises its share by being easy for an engine to verify and by catching the Google search that follows an AI mention. Neither step guarantees a position, and claims on condition pages need clinical and legal review before they ship. The sequence ClinicAds runs for a DTC brand:
- 1. Build the 24-prompt national set across conditions, price questions, legitimacy checks, and brand-versus-brand comparisons
- 2. Run each prompt 5 times in ChatGPT and Perplexity and log mention share, first-named share, and cited URLs
- 3. Search the brand name and brand name plus reviews in Google, and note whether the brand's own pages or third-party review and affiliate pages hold the top results
- 4. Run a brand-name search campaign so a competitor bid cannot intercept the searches an AI mention creates
- 5. Publish condition pages that state eligibility, price, prescriber licensing by state, and pharmacy partner in the first 60 words
- 6. Add a how-did-you-hear-about-us intake option for AI assistants and track those patients' month-3 retention separately
When is the first slot the wrong target?
The first slot is the wrong target for a telehealth brand whose payback is already broken. If paid CAC sits above $250 or payback runs past 3 months, extra patients from AI answers inherit the same retention problem, and churn work comes first. It is also the wrong target for a brand that does not own its brand-name results page, because the measured lift arrives there and leaks to whoever ranks or bids on it.
For a brand with healthy unit economics, 3:1 or better LTV to CAC, and $10,000 to $50,000 a month in paid spend, mention share is a slow second channel worth measuring quarterly. Treat first-named share as a diagnostic, not a goal.
Is being named first in ChatGPT worth more than being named third?
Not on current evidence. A June 2026 Scrunch AI preprint found the off-platform effect of an AI recommendation was flat across mention order. No telehealth-specific study of answer position was found.
Why does brand search matter for AI visibility?
In the Scrunch AI study, an AI recommendation raised same-name Google searches by 4.3 percentage points, more than direct site visits. A brand that does not hold its own name in Google loses part of that value.
How many AI answers does a telehealth brand need to measure?
ClinicAds uses about 120 answers per engine: 24 national prompts, each run 5 times. Smaller samples swing too much from run to run to act on.
Does ClinicAds guarantee a position in AI answers?
No. AI answers vary by engine and by run, and ClinicAds reports CAC, LTV, and payback figures as agency averages, not guarantees.