Measure waiting time from the right events

A number that has been called has not necessarily finished being served. This small distinction determines what queue records can tell you. If the system only stores ticket collection and call times, those records cannot establish how long service at the desk took.

Start with the question you want to answer. Are visitors waiting too long for the first call? Does document checking take a long time? Do transfers add waiting time? Each question needs different recording points.

Define the start and finish

To measure the wait for the first call, define the queue entry time and the first call time. If a visitor booked yesterday but arrived today, the reservation creation time is not the right starting point for waiting in the service area.

Measuring service duration requires a service start and finish record. A “call” action may happen before the visitor reaches the desk. Treating a call as the start of service therefore needs to be labelled as an approximation, not assumed to be the same event.

For journeys across counters, decide whether a report covers the whole journey or just one service. Both are useful as long as the meaning of the numbers stays clear.

Keep more than summary figures

An average can be one summary, but operational review needs to consider the spread of cases. Group them by service and arrival time. Look at the longest waits too, then trace observable causes.

A long record might represent a busy queue, an absent visitor, or a status someone forgot to update. Without checking, all three look like the same problem. Add the necessary context so reports do not simply penalise staff for incomplete records.

Compare like with like

To assess a change after introducing the system, record the previous process using the same definitions. Consider service types, opening hours, active counters, and procedure changes. Comparing a mass-registration opening day with a normal service day can easily mislead.

In Antrean, completion tracking and duration reporting should be treated as separate development work until their implementation is verified. Existing call data can be a starting point, but should not immediately be labelled “service completion time”.

Start the first report with definitions for every column and a few manually checked sample journeys. If staff can explain where the numbers come from, the report has a basis for decisions. Only then agree improvement targets and how to measure them. A target is something to test; it is not yet a result suitable for a marketing claim.