Service Operating Model All articles
Service Improvement

The Appointment That Didn't Happen: Reading No-Shows as Evidence, Not Excuses

Service Operating Model
The Appointment That Didn't Happen: Reading No-Shows as Evidence, Not Excuses

Photo: Claudio Centonze, European Commission, CC BY 4.0, via Wikimedia Commons

A missed appointment is almost always recorded as something the customer did. The slot is marked as not attended, the visit is closed, the capacity is written off, and the conversation moves on to how many reminders were sent. In a dental practice, a broadband installation round or a council repair service, the same label does the same work: it locates the problem outside the operating model.

That label is a decision, and it is usually the wrong one. A no-show is a recorded outcome of a booking rule, a reminder design, an access assumption and a capacity plan that were all set by someone, at some point, without the customer in the room. Read collectively, missed appointments describe the shape of the model more accurately than any satisfaction score.

Start with the denominator you actually have

Most systems count attendances well and count offers badly. Before any analysis, establish a funnel: slots offered, slots booked, slots confirmed, slots attended. If you cannot produce the first number, you cannot compute a meaningful non-attendance rate, and any comparison between services is quietly meaningless.

A useful discipline is to publish the rate three ways — against booked slots, against offered slots, and against offered slots weighted by lead time. The three rarely move together, and the differences are where the operating questions live.

Four cuts that repay the effort

Lead time. Compare appointments booked more than four weeks out with those booked within seventy-two hours. A steep gradient usually indicates that the booking window is set for the convenience of the rota rather than the circumstances of the customer. Long lead times are not inherently wrong, but they carry a maintenance obligation: confirmation, re-confirmation and a painless way to move the slot.

Channel of origin. A slot booked by a referral form behaves differently from one booked by phone, in person, or through an app. Where the contact centre produces a materially lower attendance rate than digital booking, the difference is often a data quality problem wearing a behavioural costume — wrong mobile numbers, mistyped postcodes, appointments made on someone else's behalf.

Time and place. Cluster non-attendance by site, day and hour. Consistent patterns around early slots, sites with poor transport links, or locations where parking is charged are access findings, not commitment findings. In UK service settings, the practical questions are unglamorous: is there a bus, is there a lift, is there somewhere to wait, and does the appointment time assume a car the customer may not have?

Segment. First-time versus repeat, and any recorded accessibility or vulnerability flag. Be careful here. Segment analysis should be used to redesign the offer, never to build a list of customers to chase harder.

What the patterns usually say

In most audits of this kind, three findings recur. The first is that confirmation is designed as a broadcast — a text at forty-eight hours — rather than as a conversation with a route to change. The second is that there is no waitlist, so a released slot evaporates instead of moving to someone waiting. The third is that rebooking after a missed appointment is treated as an administrative loose end rather than a service commitment, and the customer is left to re-enter the queue at the back.

None of these are customer behaviours. All three are model choices.

Three levers, and one you should resist

The reflex response to a poor attendance rate is another reminder campaign. It is cheap, visible and almost always ineffective beyond a single percentage point, because it addresses awareness in a system where the binding constraint is friction.

Scheduling rules. Shorten the booking horizon where the service allows it, offer a choice of two slots rather than one, and make the change path as easy as the booking path.

Waitlists. A functioning waitlist converts released capacity into attended capacity. It needs an owner, a sequence and a rule for how long an offer stands.

Recovery contact. A missed appointment should trigger a contact within one working day, with a rebooking option and a clear statement of what happens next. This is part of the service, not an apology appended to it. Track the recovery rate as a service measure in its own right.

A four-week audit sequence

Week one: extract the funnel and the four cuts, and check that the definitions in the report match the definitions in the operating document. Week two: sample twenty missed appointments per service and read the case notes, not the codes. Week three: shadow the reminder process and the rebooking path end to end, as a customer would experience it. Week four: take three candidate changes to the forum that owns scheduling and agree a measurement window.

The test of a good fix

A fix is only credible if the conditions of measurement stay still. Compare attendance under the same lead-time band, the same site and the same channel. If the rate improves only because the mix shifted towards short-notice bookings, you have measured your own diary, not your service.

The appointment that did not happen is a small event with a long shadow. Read one at a time, it is noise. Read together, it is the operating model describing itself — and it is one of the few descriptions you will get for free.

All Articles

Related Articles

Reading the Runaround: What Repeated Exceptions Reveal About Your Service Model

Reading the Runaround: What Repeated Exceptions Reveal About Your Service Model

The Unpaid Shift: Auditing the Customer Labour Hidden in Your Service Model

The Unpaid Shift: Auditing the Customer Labour Hidden in Your Service Model

The Half-Life of a Help Article: Auditing Knowledge Decay Across Service Operations

The Half-Life of a Help Article: Auditing Knowledge Decay Across Service Operations