Authority appointments as a scheduling risk: why flat-rate buffer times fall short
Almost every project plan contains at least one activity that the project team does not control itself: an approval, a registration, an inspection, an appointment with a public authority. In a Gantt chart, this activity looks like any other, with a start date, duration and end date. This becomes problematic when the duration entered in the plan conveys more certainty than its basis actually justifies. The time required in the plan also depends on the availability of external appointments and on the authority’s processes. A single duration value in the Gantt chart, however, does not make these dependencies visible.

Inhalt
The usual approach and why it fails
For an authority appointment to be included in the project plan, the team needs a planning value: how much time will pass between booking and the appointment? The following example is concerned with this waiting time. How long the appointment itself takes and whether any subsequent processing by the authority is required must, where necessary, be taken into account separately.
The problem can be illustrated by two planning steps:
1. A current appointment display is adopted as the future waiting time
When planning the project, someone checks the authority’s appointment portal. The next available appointment is eight calendar days in the future. The team therefore enters a waiting time of eight days in the plan. However, it does not intend to book the appointment until a later project phase. Today’s display therefore becomes the basis for an assumption about a later point in time. Whether an appointment will also be available within eight days at that point cannot be determined from this query.
2. A flat-rate uplift is intended to safeguard the estimate
Because the team is aware that it could take longer, it adds twenty per cent to the assumed waiting time. Eight calendar days therefore become around ten. The plan now contains a waiting time with a buffer. However, the additional reserve amounts to only just under two days. The extent to which the waiting time might deviate when the appointment is booked later has still not been investigated.
Suppose that, when the appointment is booked later, the earliest available appointment is not until four weeks later. The team would then have to wait 28 calendar days from that point. However, only around ten days, including the buffer, have been planned. In this example, around 18 days are therefore missing. If subsequent work depends on the authority appointment, the delay may affect further parts of the project plan.
The problem therefore begins with the initial value itself: the eight days come from a single query. They describe availability at that point in time, but do not yet provide a sufficient basis for the expected waiting time when booking at a later date. The uplift therefore also depends on an uncertain initial value. Whether the additional time adequately covers possible delays is unclear. The crucial question is therefore: how far might the actual waiting time deviate from the estimate, and how much of that must the project plan be able to absorb?
The problem can be illustrated by two planning steps:
1. A current appointment display is adopted as the future waiting time
When planning the project, someone checks the authority’s appointment portal. The next available appointment is eight calendar days in the future. The team therefore enters a waiting time of eight days in the plan. However, it does not intend to book the appointment until a later project phase. Today’s display therefore becomes the basis for an assumption about a later point in time. Whether an appointment will also be available within eight days at that point cannot be determined from this query.
2. A flat-rate uplift is intended to safeguard the estimate
Because the team is aware that it could take longer, it adds twenty per cent to the assumed waiting time. Eight calendar days therefore become around ten. The plan now contains a waiting time with a buffer. However, the additional reserve amounts to only just under two days. The extent to which the waiting time might deviate when the appointment is booked later has still not been investigated.
Suppose that, when the appointment is booked later, the earliest available appointment is not until four weeks later. The team would then have to wait 28 calendar days from that point. However, only around ten days, including the buffer, have been planned. In this example, around 18 days are therefore missing. If subsequent work depends on the authority appointment, the delay may affect further parts of the project plan.
The problem therefore begins with the initial value itself: the eight days come from a single query. They describe availability at that point in time, but do not yet provide a sufficient basis for the expected waiting time when booking at a later date. The uplift therefore also depends on an uncertain initial value. Whether the additional time adequately covers possible delays is unclear. The crucial question is therefore: how far might the actual waiting time deviate from the estimate, and how much of that must the project plan be able to absorb?
What a series of measurements shows
The appointment portals of German vehicle registration offices are checked several times a day to determine when an appointment is available for a defined reference procedure. For each office, the median regular lead time over the previous 28 days, the observed range, the number of measurements and the time of the most recent check are published. Only comparable offices with a sufficient number of recent measurements are included in the ranking and the overall median. Short-term outliers and erroneous measurements are not included in the key figures. If a portal explicitly reports that no appointments can be booked, this is shown separately and without stating a number of days. The values describe observed appointment availability. They do not measure an authority’s workload and do not guarantee the waiting time for a booking made at a later date. By 12 September 2026, 33,364 individual measurements had been carried out at 237 comparable offices.
For other types of authority procedures, the project team should separately examine how external appointment availability and any subsequent processing affect the time required.
Three observations are of interest for planning:
1. The median across all offices is five days, but 17.7 per cent of offices are above 14 days, and the highest value is 53 days. Anyone planning with a nationwide benchmark is planning for an authority they are not actually dealing with.
2. Even within the same vehicle registration office, the observed appointment lead times can fluctuate considerably. The following examples show how little a single query says about this development: one office ranged between 13 and 80 days during the measurement period, another between 0 and 66 days, and a third between 1 and 53 days. A look at the portal is therefore not a forecast, but a sample with a sample size of one.
3. Appointment availability can change as new slots are released and previously booked appointments become available again. It may therefore be worthwhile checking repeatedly whether new booking options have become available.
To illustrate this, consider the following example:
A fleet project provides for the handover of twelve vehicles on a specified date, with registration being the final step beforehand. Planning is based on the value shown in the portal, eight days, plus a twenty per cent buffer, i.e. around ten days. If the lead time rises to 30 days that week, the shortfall is not two days, but three weeks, and for all twelve vehicles at the same time because they share the same queue. This is precisely the costly part: a shared risk buffer at the end of the chain would have covered the situation; twelve individual percentage buffers do not. If the handover is additionally decoupled, for example through staggered delivery, the worst-case impact is measured in days rather than weeks.
For other types of authority procedures, the project team should separately examine how external appointment availability and any subsequent processing affect the time required.
Three observations are of interest for planning:
1. The median across all offices is five days, but 17.7 per cent of offices are above 14 days, and the highest value is 53 days. Anyone planning with a nationwide benchmark is planning for an authority they are not actually dealing with.
2. Even within the same vehicle registration office, the observed appointment lead times can fluctuate considerably. The following examples show how little a single query says about this development: one office ranged between 13 and 80 days during the measurement period, another between 0 and 66 days, and a third between 1 and 53 days. A look at the portal is therefore not a forecast, but a sample with a sample size of one.
3. Appointment availability can change as new slots are released and previously booked appointments become available again. It may therefore be worthwhile checking repeatedly whether new booking options have become available.
To illustrate this, consider the following example:
A fleet project provides for the handover of twelve vehicles on a specified date, with registration being the final step beforehand. Planning is based on the value shown in the portal, eight days, plus a twenty per cent buffer, i.e. around ten days. If the lead time rises to 30 days that week, the shortfall is not two days, but three weeks, and for all twelve vehicles at the same time because they share the same queue. This is precisely the costly part: a shared risk buffer at the end of the chain would have covered the situation; twelve individual percentage buffers do not. If the handover is additionally decoupled, for example through staggered delivery, the worst-case impact is measured in days rather than weeks.
Three consequences for planning
The observations result in three tasks for the project team:
1. Clarify the process actually required
First, it must be established which service is required, which authority is responsible and which booking or processing route applies to the specific procedure. The waiting time for an appointment, the duration of the appointment itself and any subsequent processing should be considered separately. A portal value is useful only if it corresponds to the required procedure.
2. Take several possible developments into account
A single query provides current information. For an appointment that will be required at a later point in time, the team should additionally examine how availability develops over time. Are short-notice appointments regularly offered? Do longer lead times occur? Does the portal temporarily have no bookable appointments? On this basis, a normal scenario and an adverse scenario can be described plausibly. Both should be supported by observations or transparent assumptions. If only limited data are available, this uncertainty must remain visible. A series of observations carried out by the team itself can provide initial indications, but should not delay an early booking where one is already possible.
3. Align the reserve with the consequences for the project
The initial estimate and the additional reserve should be shown separately. This makes it possible to assess which delay the buffer is intended to absorb and which activities would be affected if it proved insufficient.
If several activities depend on the same appointment availability, this shared dependency must be explicitly taken into account. A time reserve shown before an important milestone can be helpful, provided its size is appropriate to the delays being considered. In addition, the team should assess whether preparations can begin earlier or individual results can be made available independently of one another.
1. Clarify the process actually required
First, it must be established which service is required, which authority is responsible and which booking or processing route applies to the specific procedure. The waiting time for an appointment, the duration of the appointment itself and any subsequent processing should be considered separately. A portal value is useful only if it corresponds to the required procedure.
2. Take several possible developments into account
A single query provides current information. For an appointment that will be required at a later point in time, the team should additionally examine how availability develops over time. Are short-notice appointments regularly offered? Do longer lead times occur? Does the portal temporarily have no bookable appointments? On this basis, a normal scenario and an adverse scenario can be described plausibly. Both should be supported by observations or transparent assumptions. If only limited data are available, this uncertainty must remain visible. A series of observations carried out by the team itself can provide initial indications, but should not delay an early booking where one is already possible.
3. Align the reserve with the consequences for the project
The initial estimate and the additional reserve should be shown separately. This makes it possible to assess which delay the buffer is intended to absorb and which activities would be affected if it proved insufficient.
If several activities depend on the same appointment availability, this shared dependency must be explicitly taken into account. A time reserve shown before an important milestone can be helpful, provided its size is appropriate to the delays being considered. In addition, the team should assess whether preparations can begin earlier or individual results can be made available independently of one another.
What this means for the project team
The specific risk of delay should be recorded in the risk register, for example: “The required authority appointment is only available at a later point than the project schedule allows.” This should include the possible impacts, a responsible person and prepared countermeasures.
These measures must be appropriate to the procedure. They include: completing documents at an early stage, booking a possible appointment in good time, clarifying permissible alternative processing routes with the responsible authority, or monitoring appointments that become available at short notice. Equally important is the question of which subsequent activities can continue independently.
A robust project plan makes three things visible: the assumed time requirement, the justified reserve and the courses of action available in the event of a delay. In this way, an uncertain appointment assumption becomes a dependency that the project team can monitor and manage in a targeted manner.
These measures must be appropriate to the procedure. They include: completing documents at an early stage, booking a possible appointment in good time, clarifying permissible alternative processing routes with the responsible authority, or monitoring appointments that become available at short notice. Equally important is the question of which subsequent activities can continue independently.
A robust project plan makes three things visible: the assumed time requirement, the justified reserve and the courses of action available in the event of a delay. In this way, an uncertain appointment assumption becomes a dependency that the project team can monitor and manage in a targeted manner.

Author: IAPM guest - The author operates a digital vehicle registration service and publishes ongoing measurements of appointment lead times at German vehicle registration offices.
Key words: project management, schedule risk, risk management, project planning
