Appointment gaps can appear even when every patient attends. A new booking may split a free hour into intervals too short for another service. Preferring slots next to existing appointments can help, once treatment duration, clinician availability and required breaks have been checked.

One free hour, two booking choices

Fictional example: a clinician works from 09:00 to 12:00. Existing appointments occupy 09:00–10:00 and 11:00–12:00. A new consultation takes 30 minutes with no extra buffer. In this example, the shortest appointment type also takes 30 minutes.

  • Booking at 10:00 leaves an uninterrupted 30-minute slot from 10:30 to 11:00, enough for another consultation.
  • Booking at 10:15 leaves two 15-minute intervals. Neither can accommodate another appointment.
  • If the patient can only attend at 10:15, practice rules determine the decision. Calendar utilisation is not the only consideration.

How should slot suggestions be ordered?

  • Check eligibility first: service, clinician, location, full duration and required room or equipment.
  • Respect buffers: room preparation, staff breaks and any other protected time. Do not shorten them to improve utilisation.
  • Then order suitable options: prefer times next to existing appointments when they work for the patient.
  • Define exceptions: decide who handles requests when no suggested time works or a patient reports an urgent need. Clinical assessment belongs to staff.

Finding available slots and choosing between them are separate tasks. ElevenLabs Reception.ai scheduling documentation describes availability checks involving staff, resources and booking rules. Calendar access alone does not establish that a tool prioritises adjacent appointments.

What should a supplier demonstrate?

  • Create a test schedule with a free hour between appointments and request a 30-minute consultation.
  • Repeat with a longer service, a buffer and a patient who can attend at only one particular time.
  • Have reception take the proposed slot during the conversation. Availability should be checked again before booking.
  • Inspect the actual PMS: determine whether there is a confirmed appointment or a request awaiting approval.
  • Check that existing appointments have stayed in place. Improving suggestions does not authorise moving other patients.

How can you measure improvement?

Over comparable weeks, count free minutes in the bookable schedule and minutes in intervals too short for any permitted appointment including its buffer. Exclude planned staff breaks. Separate gaps created during booking from those caused by cancellations.

Alongside utilisation, track booked and attended appointments and requests declined because no convenient time was available. Do not judge the rule solely by a full calendar. Compare the same clinicians, working hours and service mix; a change alone does not prove an AI effect.

FAQ

Can every voicebot avoid appointment gaps?

Booking capability alone does not establish this. Ask about slot-selection rules, integration scope and a test using your schedule. A fluent conversation is not proof that the booking works.

Can AI move other patients automatically?

Preferring suitable free slots does not authorise changes to existing appointments. Rescheduling needs a separate agreed process and confirmation with the patient.

Should every free minute be filled?

No. Buffers, breaks and time reserved for urgent needs may be intentionally unavailable. Define them before measuring gaps so the metric does not encourage removing protected time.

How do I check this scenario with receptionOS?

Bring your PMS name, service durations and an example schedule without patient data to the demonstration. Ask whether this specific rule is available for your integration. Check supported operations on the receptionOS integrations page, prepare rules with the custom scenarios guide and use the voicebot acceptance tests.