Skip to content
← All case studies
Client delivery UK employer (sector withheld)

Payroll preparation: from 48 hours to 40 minutes

A UK employer with 250–1,000 staff · Published

Before
48 hours
After
40 minutes
Payroll preparation turnaround: what was measured

Elapsed turnaround per payroll cycle: the time from the cycle's inputs being due to a validated, payroll-ready input set being available. It is calendar time, including waiting and chasing, not the hours people spent actively working.

The situation

Before each payroll run, the client’s team had to assemble a complete, correct set of inputs: hours worked, changes to employment records, starters and leavers, and the adjustments that make every payroll cycle slightly different from the last. Those inputs lived in several places, arrived at different times and were checked by people who knew, from experience, what a wrong figure looked like.

The payroll run itself was not the problem. The preparation was. Most of the 48 hours between inputs being due and the payroll being ready to process was spent waiting for information, chasing what was missing, re-keying data between systems and finding errors late enough that fixing them meant going back to the source.

The constraint

Payroll is unforgiving. A wrong figure reaches a real person’s pay, and errors are expensive to reverse and damaging to trust. Any change had to make the process faster and at least as reliable, keep the people responsible for payroll in control of the final result, and leave the payroll system of record untouched.

What we changed

The work focused on collection and validation, not on the payroll calculation:

  • Automated input collection. Timesheet and HR record changes are gathered automatically from the systems that hold them, instead of being requested, exported and assembled by hand.
  • Rule-based validation at the start. The checks the team already performed from experience were written down and encoded as validation rules, so missing, inconsistent or out-of-range inputs are flagged as soon as they arrive rather than discovered at the end.
  • An exception list instead of a full review. People now review the specific items the rules flag, with the context needed to resolve them, instead of re-checking every input line by line.

The team responsible for payroll kept final sign-off. Nothing in the change decides pay; it removes the collection and checking work that stood between the inputs and that decision.

The result

Elapsed turnaround for payroll preparation fell from 48 hours to 40 minutes per cycle.

Most of the gain came from removing waiting and rework rather than from doing the remaining work faster. When errors surface at input time, they are fixed by the person who supplied the input, while the context is fresh, instead of being traced back days later.

What this does and does not show

This figure is elapsed turnaround, measured per cycle and compared before and after the change. It is not a measure of total labour saved, of payroll accuracy, or of cost, which are different quantities. It describes one employer’s process with its particular systems and rules; a different organisation’s starting point would produce a different result. The client has approved this anonymised account.

Have a workflow like this?

Tell us how often it runs, which systems it touches and where the time goes. We will suggest how to measure it before anything changes.

Describe your workflow