For twenty years, being found meant ranking on a search page a human then scrolled. That human is increasingly delegating the search. A prospective client no longer always types “yoga studio near me” and browses ten blue links — sometimes they ask an AI assistant to find a well-reviewed practitioner with Thursday-evening availability and, more and more, to go ahead and book it. When the person doing the looking is a machine acting for your client, the question is no longer only “does your website look good,” but “can an AI assistant read, understand, and act on what you offer?”
A new front door you did not design for
Most wellness businesses are, right now, invisible to this channel. Their services live as prose on a page and as options buried three clicks deep in a booking widget — legible to a person with patience, opaque to an assistant trying to match a request to a bookable thing. If an AI cannot parse that you offer a 60-minute prenatal massage on Thursdays at a given price, it cannot surface you as an answer, and it certainly cannot complete the booking. Being human-readable is no longer the same as being discoverable, because a growing share of the discovery is happening on the other side of a machine.
The honest framing matters here, so let us be plain: this is an early and growing channel, not a settled one. The share of wellness bookings that originate from an AI assistant today is small, and anyone who quotes you a precise figure is guessing. But the direction is not ambiguous, and the businesses that become machine-readable early are the ones an assistant can choose when the volume arrives. Discoverability is a position you take before the channel matures, not after.
Structured services an assistant can find and choose
Becoming discoverable to AI assistants is less about marketing copy and more about structure. An assistant chooses what it can understand: services expressed as clear, structured, machine-readable objects — this offering, this duration, this price, this availability — rather than paragraphs it has to interpret. The emerging standard for this is MCP (the Model Context Protocol), which lets an AI assistant discover a business’s capabilities and act on them directly. A business that exposes its services over MCP is not just online; it is callable — present in the set of options an assistant can actually pick from and book.
This is where owning your platform, rather than renting fragments of it, becomes a discovery advantage. VBWD ships MCP natively as one of its plugins over an agnostic core, so an assistant can discover and book or buy against your services directly — and because booking, catalogue, and content already live in the same self-hosted system, what the assistant sees is the same truth your clients see, not a stale export. You enable it, like any plugin, without a restart. VBWD sketches how a small operator runs this class of infrastructure in its write-up on running an enterprise-grade stack as a small studio, and the module map is on the features overview.
The scale point cuts the same way as everything else here: because the infrastructure does the heavy lifting, a single-location studio can present itself to AI assistants exactly as a large multi-location group would, from one backend serving web, iPhone, and Android. Being discoverable to machines is not reserved for businesses with an engineering department.
The honest caveat
Two caveats deserve to stand next to the optimism. First, as said, the AI-assistant discovery channel is genuinely early — it is a bet on where client behaviour is heading, and you should size the investment as a bet, not a certainty. Second, the platform question carries the usual trade: VBWD is younger than the twenty-year-old booking tools it can replace, and it has fewer accumulated edge-case behaviours to show for it. What it offers instead is modern architecture — including native MCP — plus speed, auditability, and data sovereignty, with your client records staying inside your own jurisdiction rather than on infrastructure you do not govern. For a wellness business that wants to be found by the assistants its clients are starting to use, that forward-leaning architecture is often the better trade; for one whose clientele will book by phone for years yet, the urgency is lower. It is source-available under BSL 1.1 — free for commercial use while annual VBWD-attributable sales stay below the value of 6.7 BTC per year — so you can experiment before you commit.
If your practice is wrestling with any of this — the sense that clients are starting to search through an assistant you are invisible to, the services no machine can read, the booking system that fights both humans and software — the useful next step is concrete: see your services made discoverable and callable for your own business. Request an enterprise installation and bring the numbers you want to improve.
