OpenAI's Anima page reports that clinics cut call waits by 75% using GPT-6 and GPT-Live, as stated in its title. The page summary says the models give care teams more capacity to respond to patients (OpenAI).
For an appointment-based business, that is a service-speed result—not proof that more clients return. Faster answers might help a first-time caller book an initial visit, or make it easier for a regular to reschedule before giving up. Whether either change leads to a completed follow-up visit is a separate question to measure.
Why faster responses could matter for retention
Think of the call as one step in the client journey, not the retention outcome itself. A person may reach out because they are considering a first appointment, want to rebook, need to change a booking or have a question before returning. A quicker response could reduce friction at those moments. It could also have little effect if the caller's reason for reaching out is unrelated to booking.
That makes Anima's reported result a useful hypothesis for your own business, rather than a result to assume will transfer. The relevant question is not simply whether an AI tool answers quickly. It is whether a change in response experience helps more clients complete the next visit.
Start with a clear outcome. For first-time clients, track the share who complete a second visit within a consistent period after their first appointment. For existing clients, track completed rebookings or completed visits after a reschedule. A booking made is useful to observe, but it is not the same as a visit completed.
Measure the return, not just the response
Before adopting a new response process, decide which client group and follow-up outcome matter most. Keep the question narrow: for example, whether first-visit follow-up leads to more completed second visits, or whether regulars who contact you to reschedule return for their next appointment.
Then review the steps between contact and return. If your systems allow it, track when a client first called, whether the call was answered or routed to another channel, whether an appointment was booked and whether that visit was completed. Use only data you are permitted to connect. If call and booking records cannot be linked reliably, start with a simpler measure you can trust rather than treating an incomplete count as a retention result.
Use attribution experiments—tests that estimate whether a change contributed to measurable lift—to assess the effect. Do not judge the effort by response speed alone. Pair that operational measure with the client outcome you care about, such as completed second visits or completed rebookings. If answers become faster but return visits do not change, the tool may still help service operations, but the evidence would not show a retention lift.
Keep client data appropriate
You do not need to send client details to advertising platforms to assess rebooking. Use the booking and service records your business is permitted to analyze, and keep the measurement focused on the outcome. For med spas, IV and wellness clinics, check whether HIPAA or other privacy rules apply. Do not share client health or treatment details with ad platforms or general marketing tools.
What to do
- Choose one retention outcome, such as completed second visits or completed rebookings.
- Write down which client group and follow-up period you will assess before reviewing results.
- Check whether your call and booking data can be connected reliably and appropriately.
- Use an attribution experiment to assess whether faster responses contribute to a change in that outcome.
- Review response speed and completed visits together, not as interchangeable measures.
OpenAI's Anima story makes faster patient response a concrete AI use case. For your business, the next step is to find out whether that improvement reaches beyond the phone and helps clients come back. Vetraeus's Value Attribution service uses attribution experiments to measure real marketing lift, including the effect of retention work.
- applied-ai
- client-retention
- rebooking
- attribution
- appointment-businesses
Alan Jordan, founder and principal solutions architect. Data engineering, data science and business analytics for appointment-based businesses.
