When it is the right approach
Ad hoc handling means dealing with something as it arises rather than through a defined process, and there are situations where that is correct.
- Genuinely novel circumstances, where no process exists because nothing like it has happened.
- Situations where speed matters more than consistency, such as an urgent safety issue.
- Low-frequency, low-stakes matters where building a process would cost more than handling each case.
- Pilots and experiments, where the point is to learn before deciding what the process should be.
Building a process for everything is its own failure. An organisation with a documented procedure for every situation has usually stopped being able to respond to anything unusual, and its people spend their time on compliance with process rather than on the work.
The judgement is about frequency and stakes, not about tidiness.
When it becomes a problem
Ad hoc stops being appropriate at roughly the third occurrence. By then it is not a novel situation, it is a recurring one being handled without a method.
| Symptom | What it indicates |
|---|---|
| The same question is answered differently by different people | A decision that needs a rule |
| Handling depends on who is asked | Consistency is already lost |
| Nobody can explain a past decision | No record was kept, because there was no process to keep one |
| One person is the only one who knows how this is done | A dependency, not a process |
| The exception is more common than the rule | The rule is wrong |
The last row is worth acting on rather than working around. Where most cases are handled as exceptions, the standing process describes a situation that rarely occurs, and the fix is to change the rule rather than to keep granting exceptions.
The particular risk with people decisions
In an HR context, ad hoc handling carries a specific cost beyond inefficiency, which is inconsistency between people.
An ad hoc decision about one person's leave, pay, working pattern or disciplinary outcome creates a precedent that nobody recorded and that the next decision-maker does not know about. The result is two employees in comparable situations treated differently, with no explanation available.
That is the failure that produces grievances, and it is difficult to defend precisely because there is no record of the reasoning. An inconsistency with a documented reason is arguable; an inconsistency nobody can account for is not.
- Record ad hoc decisions about people and the reasoning, even where no process exists.
- Check what was decided in comparable cases before deciding a new one.
- Where a decision will obviously recur, decide the rule at the same time rather than after the third grievance.
- Keep decisions about pay, leave and conduct out of ad hoc handling wherever possible, since these are the ones most compared.
The first is the cheap control. Recording a one-off decision costs a few minutes and preserves the ability to explain it, which is the entire difference when it is questioned a year later.
The cost of formalising
The instinct after a bad ad hoc outcome is to build a process, and that instinct is right less often than it feels.
A process has ongoing costs: it has to be maintained, communicated, trained, applied to cases it fits badly, and revised when circumstances change. For something that happens twice a year it is frequently not worth it.
The middle option is generally better and is under-used: a stated principle rather than a procedure. Deciding that requests of a certain kind are approved by a named role, using a stated consideration, gives consistency without a documented workflow.
The decision is not between ad hoc and full process. It is about how much structure the frequency and the stakes actually justify, and for most recurring HR situations the answer is a rule and a record rather than a procedure.
Ad hoc work and workload
There is a second sense worth mentioning, which is ad hoc work: unplanned tasks arriving without notice.
In HR this is a large share of the actual job, and it is systematically excluded from planning. A team resourced against its planned work will be consistently overloaded, because the unplanned work is real and recurring even though its individual items are unpredictable.
Measuring the proportion of time going to unplanned work is worth doing once. It is usually much higher than assumed, and it explains why planned initiatives slip repeatedly without anyone being able to point at what consumed the time.
The response is to resource for it explicitly rather than treating each instance as an interruption to be absorbed.
Frequently asked questions
What does ad hoc mean in an HR context?
Something done as the need arises rather than through a standing process: a one-off decision, an unplanned task, or an arrangement made for a particular case.
When is ad hoc handling appropriate?
For genuinely novel situations, where speed matters more than consistency, for low-frequency and low-stakes matters where a process would cost more than the cases, and for pilots where the point is to learn before deciding the rule.
When does ad hoc become a problem?
Around the third occurrence. By then it is a recurring situation handled without a method, and the symptoms are the same question answered differently by different people and nobody being able to explain a past decision.
Why is ad hoc handling riskier for people decisions?
Because it creates inconsistency between employees with no record of the reasoning. An inconsistency with a documented reason is arguable; one nobody can account for is not, and that is what produces grievances.
Should every recurring situation become a process?
No. Processes have to be maintained, communicated and applied to cases they fit badly. For something happening twice a year, a stated principle and a named decision-maker usually gives the consistency without the overhead of a procedure.
How Engage keeps decisions traceable
Engage records decisions and their stated reasons against the employee record even where no standing workflow exists, so a one-off arrangement can be explained a year later. Because comparable cases are reportable, the question of what was decided last time is answerable before the next decision rather than after a grievance.
See employee records in Engage