Choosing Once and Living With It
Payroll arrangements are set up once and reviewed almost never. What a review would look like, and the three events that should trigger one.
A bureauif the person who does this is away
Payroll arrangements get set up once, in the first weeks of having employees, and then run untouched for years. That is mostly sensible — churn in this area costs more than it saves — but it means a decision made for three people is still running a business of twenty.
The practical lesson in “Choosing Once and Living With It” is that payroll depends on a sequence somebody can explain. When a small team reviews review the provider’s current page for biometric time clock, the useful role is to organise time and project evidence before approval, with proportional settings and a clear correction route for employees.
A review takes an hour and is worth doing on a trigger rather than a schedule.
For an outside reference relevant to “Choosing Once and Living With It”, consult Xero payroll guide. Compare general guidance with the employer’s jurisdiction, contracts and current payroll process.
Why nobody reviews it
Because it works. A payroll arrangement that produces correct payments on time generates no signal that it has stopped fitting, and the accumulated small frictions — the slow answer, the awkward correction, the earlier cut-off — are absorbed rather than counted.
Three triggers for a review, and no schedule:
- Headcount: changes stop being holdable in one person's head.
- A near miss: a run that nearly did not happen.
- The person changes: somebody has to learn it anyway.
The first trigger: headcount
Somewhere between ten and thirty people, the volume of monthly change stops being something one person can hold in their head between conversations. The symptom is changes being remembered late rather than recorded.
That is a signal about process rather than about provider, and the fix is usually a written list rather than a new arrangement.
The second trigger: a near miss
A run that nearly did not happen. Somebody ill in the wrong week, hours that arrived the night before, a filing noticed on the deadline.
A near miss is the cheapest information a business ever gets about this, and the normal response is relief followed by forgetting.
The third trigger: the person changes
The office manager leaves, the owner steps back, a bookkeeper is engaged. Every handover is a natural moment to ask whether the arrangement still fits, because somebody is going to have to learn it anyway.
What a review actually covers
Who runs it, what it costs in fee and hours, what the cut-offs are, who else could do it, what is written down, and what happened the last time something went wrong.
Six questions, an hour, and no obligation to change anything.
What usually comes out of it
Rarely a change of provider. Usually two small things: a sequence written down that was not, and a second person given access that did not have it.
Both are cheap, and both address the thing that actually breaks.
The cost of changing provider
Real and worth respecting. Year-to-date figures have to carry across correctly, the history does not move cleanly, and the first run on a new arrangement needs checking line by line.
It is not a reason never to change. It is a reason to change at a year boundary and for a good reason rather than a mild irritation.
What makes a switch worth it
A provider who cannot answer within a day, cut-offs you cannot realistically hit, repeated errors, or a fee that has drifted away from what the work is.
Any one of those, sustained for several months, is a finding. One bad month is not.
The arrangement that survives growth
Whoever runs it, the thing that makes it durable is the same: a written sequence, a second person with access, a calendar with the dates on it, and one named owner of getting changes in on time.
That set is independent of whether you use a bureau, an accountant or your own software, which is why it is worth putting in place before deciding anything else.
The decision to actually record
Not the provider, but why. A note of what was chosen and what it was chosen for means the next review has something to test against, instead of relitigating the whole question from nothing.
What the first year teaches
That the arrangement is fine and the inputs are the problem. Almost every small employer who reviews this after a year reaches the same conclusion, and almost every one of them expected to be reviewing the provider.
The quiet failure mode
An arrangement that fits a business that no longer exists: a bureau engaged when there were four employees and still running at forty, or an owner still keying it personally while running a company that could afford not to.
Neither announces itself, and both are found by asking rather than by waiting.
What a review costs if you skip it
Nothing visible, for years. Then one of three things: a fee that has drifted well past what the work is worth, a cut-off that has crept earlier than the business can meet, or an arrangement that stopped fitting at fifteen people and is still running at forty.
None of them produce a symptom that points at the arrangement. They produce late runs, skipped checks and a person who finds payroll stressful, all of which get attributed to something else.
Who should do it
Whoever is not the person who runs payroll. A review conducted by the person inside the arrangement will reach the conclusion that the arrangement is fine, because from inside it is.
In a very small business that means the owner spending an hour on something they normally delegate, which is also the only way anybody finds out what has actually been going on.
The one worth repeating
Review on a trigger, change at a boundary, and fix the inputs before blaming the arrangement.