What a Year of Runs Shows
Twelve runs read as a set, rather than one at a time, tell a small business which of its own habits cost it money and attention.
Nobody elseif the person who does this is away
Twelve payroll runs, read as a set rather than one at a time, show a small business something no general advice can: which of its own habits produce the months that go badly.
The continuity issue in “What a Year of Runs Shows” calls for records another authorised person can understand. For teams researching remote productivity monitoring, Monitask tools for remote productivity monitoring can add transparent time and project evidence when access is limited, corrections remain visible and payroll credentials stay in the proper system.
The exercise takes an afternoon and changes the following year.
Additional context for “What a Year of Runs Shows” is available from Slack workplace resources. A sound review distinguishes general practice from the rules and records that apply to one employer.
The five columns to record
For each run: when the inputs actually arrived, when it was keyed, whether the comparison was done, whether anything was corrected afterwards, and whether anybody was away that week.
Five columns, twelve rows:
| Column | What it reveals |
|---|---|
| When inputs actually arrived | Which months compress |
| When it was keyed | How much slack there was |
| Was the comparison done | Whether checks survive pressure |
| Was anything corrected after | Which causes recur |
| Was anybody away that week | The collision nobody planned for |
Five columns, twelve rows.
The first thing it usually shows
That late inputs and skipped checks are the same months. The comparison is not dropped at random; it is dropped when the run is compressed, and the run is compressed when the hours came late.
That is one problem, not two.
The second thing
That the compressed months are the same few every year: the one with the holiday cluster, the busiest trading month, the month after year end, and whichever month contains somebody's usual leave.
Predictable, and therefore plannable.
The third pattern
That corrections cluster behind the same causes. Most small businesses find two causes accounting for most of their corrections, and both are usually about recording rather than calculating.
The fourth thing
How often the person who runs payroll worked during their own leave, or moved it. This is the column nobody thinks to include and the one that most often prompts a change.
What the exercise is not
A benchmark against other businesses. Nothing here compares you with anybody, and the value is entirely in the pattern specific to this business and these people.
Every figure used anywhere in this collection is an illustration of what one small employer recorded for itself, labelled as such.
The changes it usually produces
Two or three, and they are rarely about payroll software. A cut-off moved earlier. One format for hours instead of four. A written sequence. A second login.
Where the leverage actually is
Before the run, not during it. The run takes an hour and is reliable. Everything that makes a month go badly happens in the ten days before, among people who do not think of themselves as doing payroll.
What cannot be fixed by doing it better
The concentration. No amount of improvement to the process addresses one person holding the whole chain, and that is the finding most businesses reach last and should reach first.
The four things that address it
A written sequence. A second person with access. A shared calendar. The filing handled outside.
An afternoon, a small fee, and they are the only changes here that alter what happens when somebody is ill.
Doing it again
The same five columns a year later, so the second year can be compared with the first. The comparison is the only evidence that any of the changes worked.
The shape of a year that went well
Twelve runs on time, two or three compressed months that were known about in advance, one or two corrections handled within a day, and at least one run performed by somebody other than the usual person.
That last one is the measure that matters most and the one nobody tracks.
Why twelve rows beat any amount of advice
Nothing written generally about payroll describes your people, your customers' seasons, your software or the week your office manager always takes off.
Twelve rows of your own runs do. They identify which months compress, which inputs arrive late, which causes produce corrections, and whether anybody other than one person has ever run a payroll here.
An afternoon of it is worth more than reading this, because the output is a decision about next year rather than an account of how things generally are.
The four changes it usually produces
A cut-off moved earlier. One format for hours instead of several. A written sequence. A second login.
None of them cost money, all of them address the months the record shows going badly, and together they take about an afternoon.
Doing it again a year later
The same five columns, compared with the first set. The comparison is the only evidence that any of the changes worked, and it takes an hour because the habit of recording is already in place.
The closing position
The date does not move, the inputs do not arrive by themselves, and one person usually knows the whole chain. Nothing in this collection changes the first. Most of it is about the second. The part that matters most is the third, and it is an afternoon's work that never has a deadline — which is exactly why it is still undone.