Should a practice open online self-scheduling?

The pitch for online self-scheduling is call deflection, and the arithmetic behind it is usually presented against calls offered — how many booking calls the practice received before, and how many after. That comparison is fine in a practice whose lines are not saturated. In a practice whose lines are saturated it is close to meaningless, because a caller who could not get through was never offered as a call in the first place. What self-scheduling does there is release suppressed demand, and the first metric to move is abandonment, not volume. Total calls can stay flat while the service improves substantially, and a practice measuring only volume will conclude the product failed.

That is worth stating plainly because it cuts both ways. Some practices will buy self-scheduling on a deflection promise and see no volume change, and be right to feel misled about the promise while being wrong about the outcome. The honest evaluation compares the share of callers who reached a person, before and after, and it needs the baseline captured before the system is switched on — which almost nobody does, because the baseline lives in a phone report nobody has opened.

The risk that actually matters is not on the phones at all. Self-scheduling writes into the same appointment book the front desk writes into, and it writes without the judgement the front desk applies. A patient selects the visit type that sounds closest to their situation, and the failure is a new-patient consultation booked into a follow-up slot, or a procedure booked with a clinician who does not perform it. That does not surface as a scheduling error. It surfaces mid-clinic, as a session that overruns, and by then the cost has been paid by every patient later in the list.

Which is why the sensible first configuration is narrow rather than open. Publish the visit types where a wrong booking is cheap and recoverable — an established patient, a single named service, a fixed duration — and keep the ambiguous ones behind a person. That is a much smaller product than the one being sold, and it captures most of the availability benefit with almost none of the integrity exposure. Widening it later is a decision you can make from data; narrowing it after a bad quarter is a decision you make from complaints.

There is a second, quieter benefit that rarely appears in the business case and is often the largest one. Self-scheduling is the only intake channel that works at two in the morning, and enquiry timing does not respect opening hours. A practice with no after-hours path is not competing badly outside its hours; it is absent, and absence is invisible in every report it keeps.

The measurement to commit to before signing anything: the share of prospective patients who reach a person, and the elapsed time from a public enquiry to a human response. Both are observable from outside the practice, with no vendor involvement and no access to the phone system, which is exactly what makes them worth holding a supplier to.