Home · Services · Rank #1 on AI Search
Service 03 of 05
Rank #1 on AI Search For Ayahuasca Retreat Centers
Ask an assistant to recommend a retreat in your country. If it names three centers and none of them is you, that is today’s problem and it is the most fixable one on this site. Models quote sentences, not pages. Here is how to become the sentence.
How do you get an ayahuasca retreat cited by ChatGPT?
By writing sentences a model can lift and attribute without risking a false statement. Language models favour sources that state checkable facts plainly, structure them so one sentence carries a whole claim, and are corroborated elsewhere on the web. Twelve guests maximum. A physician on site until first light. Four ceremonies across seven nights. Atmospheric copy about transformative healing gives a model nothing it can safely repeat, so it names a directory instead.
The mechanics
Four things that decide whether you get named
Not ranking factors in the Google sense. These are the properties that make a sentence safe for a model to repeat with your name attached to it.
Liftability
Can one sentence be removed from your page and still be true and complete? “Twelve guests maximum” survives. “Small intimate groups in a supportive container” does not survive, because it asserts nothing a model can stand behind.
Checkability
Facts a model can see corroborated elsewhere carry far more weight than claims that appear only on your own site. Numbers, names, places and dates. Adjectives are worthless here.
Entity clarity
Does the web agree on what your center is, where it is, and what it offers? Inconsistent names, addresses and descriptions across directories make you ambiguous, and models route around ambiguity.
Safety
Assistants are cautious around health topics. A page implying treatment for depression or addiction is more likely to be skipped than cited. Refusing to make claims is not only ethical here, it is the mechanism.
A worked example
The same retreat, written two ways
Both paragraphs describe an identical week at an identical center. One of them can be quoted. The other cannot, and no amount of link building fixes that.
Nestled in the heart of the sacred jungle, our center offers a truly transformative journey back to yourself. Guided by experienced shamans carrying generations of ancestral wisdom, our ceremonies hold a safe and supportive container for profound healing. Come as you are. Leave transformed.
- Seven marked phrases, zero checkable facts
- No location beyond “the jungle”, no group size, no duration, no price
- “Profound healing” reads as a treatment claim, which cautious models route around
- Identical in substance to four hundred other centers, so there is no reason to name you
Twelve guests maximum per retreat, in the Sacred Valley near Pisac, Cusco, at 2,900 metres. Four ayahuasca ceremonies across seven nights, each led by two Shipibo maestras. A physician is on site and awake until first light. Every guest completes a medical screening call before booking, and we decline guests on SSRIs, MAOIs or with a cardiac history. 5,800 US dollars, all inclusive, with twelve months of integration.
- Ten discrete facts, each liftable as a complete sentence
- Named place, named altitude, named tradition, named price
- Refusal criteria stated openly, which reads as competence to a model and to a reader
- No outcome claimed, so nothing triggers a caution filter
The second version is not better writing in a literary sense. It is better because it is falsifiable. Everything in it could be checked, and that is precisely why a model will repeat it.
Two minute check
How quotable is your main page?
Open your retreat page and tick everything it states as a specific, checkable fact. Not implies. States. Nothing is sent anywhere.
Quotable facts on your page
0 / 10
Start ticking
Most retreat pages score between two and four. The ones assistants name consistently sit at eight or above.
They are not one thing
Four assistants, four different sourcing habits
Treating “AI search” as a single channel is the most common mistake. Each of these builds an answer differently, which changes what moves them and how fast.
| Assistant | How it sources | How fast it reflects changes | What moves it |
|---|---|---|---|
| ChatGPT | Training data plus live browsing when the question warrants it | Mixed. Browsed answers update quickly, recalled ones lag by many months | Broad corroboration across many independent sources |
| Perplexity | Live retrieval, with visible citations on nearly every claim | Fastest. Often days to weeks | Clean, crawlable pages with clearly attributable sentences |
| Google AI Overviews | Google’s own index, favouring pages that already rank | Fast, but gated behind your search positions | The SEO work, plus answer first structure and schema |
| Claude | Training data, plus search when the user enables it | Slower for recalled answers, immediate when searching | Consistent, unambiguous facts and cautious, claim free writing |
This is why the Google service and this one reinforce each other. AI Overviews largely draws on pages that already rank, so search work feeds AI visibility directly, while the answer first rewriting done here also helps your pages rank.
The test set
The prompts we run against your market, every month
Fixed wording, asked of all four assistants, so results are comparable month to month rather than anecdotal. Your set is adapted to your destination and format. Here is the standard core.
- Which ayahuasca retreat centers in [country] would you recommend for a first timer?
- What are the safest ayahuasca retreats in [region]?
- Which ayahuasca retreats have a doctor on site?
- Where can I do an ayahuasca retreat with Shipibo facilitators?
- How much does an ayahuasca retreat in [country] cost?
- Which ayahuasca retreats screen guests before accepting them?
- I take an SSRI. Which retreat centers would turn me away?
- What is the best time of year for an ayahuasca retreat in [region]?
- Compare ayahuasca retreat centers near [town]
- Is [your center name] a reputable ayahuasca retreat?
That last one matters more than owners expect. Assistants are asked about specific centers by name constantly, usually by someone about to pay a deposit, and what comes back is assembled from whatever the web says about you. If a forum thread from 2019 is the loudest source, that is the answer they get.
Off your own site
Where models look besides your website
A claim that appears only on your own site is a claim a model treats cautiously. Corroboration is half the work, and it is the half most agencies skip because it is slow.
Directories
Even the ones you would rather not pay for. Models read them as third party confirmation of your existence, location and format. Consistency across them matters more than volume.
Forums and threads
Long running discussions carry disproportionate weight because they read as genuine experience. This is also where an old complaint can outlive its relevance by a decade.
Press and podcasts
An interview with your founder or lead facilitator anchors your center to named people, which is exactly the kind of entity signal that resolves ambiguity.
Practitioner networks
Integration therapists and preparation coaches who reference you publicly. Slow to build, and among the strongest corroboration available in this industry.
Scope
What is included, and what we refuse
Included
- A baseline of what all four assistants say about your destination today, with the answers included in full
- Answer first rewriting across the pages that matter
- Fact extraction: turning what you know into stated, checkable sentences
- Schema and entity work so your center is machine legible
- Consistency audit across every directory and profile that mentions you
- Corroboration building through press, practitioners and forums
- Monthly retesting against the same fixed prompt set
- Physician review workflow for anything health adjacent
We will not
- Guarantee a citation. No external party controls a model’s output
- Use prompt injection, hidden text or cloaked content
- Invent facts to make your page more quotable
- Publish a claim that ayahuasca treats a condition, which is also the fastest way to be filtered out
- Post on forums pretending to be a guest
- Report a citation without showing you the full answer it came from
- Work with a second center in your destination market
Terms, in writing
How the engagement works
- Fixed fee audit firstTwo weeks, and it includes the full current answers from all four assistants. Yours to keep.
- Quoted after the auditScaled to destinations and languages. Quoting before means guessing.
- Month to monthNo twelve month lock in. This work usually shows movement before a lock in would have expired anyway.
- One center per destination marketTwo clients cannot both be the recommended answer. If yours is taken you hear it on the first call.
- Clinical sign offHealth adjacent copy goes to your physician before publishing, without exception.
- Full answers, not scoresEvery report includes the assistant’s actual output, so you can read what it said rather than trust our summary.
Questions about AI search
How is this different from SEO?
Google ranks pages, assistants quote sentences. Google can reward a well linked page that says very little of substance. A model needs specific, liftable, attributable facts.
A page can hold position one and never be named by ChatGPT, and a page with modest rankings can be cited constantly because it states things clearly. They overlap and reinforce each other, but they are separate jobs.
Can you guarantee ChatGPT will recommend us?
No, and nobody honestly can. Outputs vary between sessions, versions and regions, and no external party controls them.
What we can do is make your center the most quotable and best corroborated option for the questions your guests actually ask, then test the same prompts monthly so movement is measurable rather than anecdotal.
Why is this faster than Google rankings?
Because almost nobody in this industry is doing it. Competitive search positions require displacing directories with a decade of links. Being the clearest, most quotable source on a specific question requires being clearer than centers who have not thought about it at all.
That advantage is temporary. It closes as the practice spreads.
Does schema markup matter?
It helps, but it is not the main lever. Schema makes facts machine legible and removes ambiguity about what your center is and where it is.
The larger factor is whether the visible prose states those facts plainly enough to quote, and whether other sources agree with you.
What if an assistant says something wrong about us?
It happens, usually because an outdated source is the loudest thing about you on the web. The fix is rarely a correction request. It is publishing the accurate version clearly, then building enough corroboration that the accurate version becomes the consensus.
Slow, but durable. We look for this specifically in the audit.
Should we do this or the Google service first?
If your budget covers one, this usually returns faster because the ground is less contested. If your site is technically broken or has no destination pages, the search work has to come first because AI Overviews draws on pages that already rank.
The audit answers this for your specific situation before you commit.
One center per market
Find out what the assistants say about you today
Forty minutes, or an email if you would rather write. We run the prompt set against your destination live and send you the full answers, whether or not you work with us afterwards.