HR Reporting Process Transformation

Business context

An HR function produced a recurring report that managers relied on. The report itself was not complicated; the process that produced it was. Each cycle depended on gathering inputs from several places, reshaping them by hand, checking totals against each other and correcting what did not agree before anything could be shared.

Challenge

The cycle took most of a working day every time it ran. Because the effort was spread across many small manual steps, nobody owned the total, and the time cost was largely invisible. The question was not “how do we build a better report?” but “where does the time actually go, and which of those steps need to exist at all?”

My contribution

• Investigated the reporting requirements with the people who produced and used the report.
• Mapped the as-is process step by step and timed it across ten conventional cycles.
• Identified the root causes of the delay rather than the symptoms.
• Designed and built the Power BI-assisted workflow that replaced the manual collation and reconciliation steps.
• Timed ten cycles of the revised process and calculated the change.

Investigation and methodology

I treated the report as a production process. Each cycle was broken into stages — gathering inputs, cleaning and reshaping, reconciling, assembling and reviewing — and each stage was timed. Recording ten cycles rather than one mattered: it gave a baseline that reflected normal variation instead of a single good or bad day.

Process mapping showed where the waiting and rework sat. Root-cause analysis then separated work that added value (review and judgement) from work that existed only because data arrived in inconsistent shapes and had to be repaired by hand each time.

Solution design and implementation

The redesign moved the repeatable work — shaping, combining and reconciling the inputs — into a modelled Power BI workflow, so it ran the same way every cycle. People kept the steps that needed judgement. The system model of the new process was documented so the steps could be followed and checked by others.

Results and business value

Measure	Conventional process	Power BI-assisted process
Cycles measured	10	10
Average cycle time	1,069.6 minutes	117.4 minutes
Change	—	−952.2 minutes per cycle (−89.02%)

The measured improvement is specific to the reporting cycle. Its value is the time released from repeated manual handling, and a more consistent output, because the reconciliation logic no longer depended on who ran the cycle.

Technical details

• Measurement design: ten timed cycles before and ten after; the averages are arithmetic means of total cycle time.
• Calculation: (1,069.6 − 117.4) ÷ 1,069.6 = 0.8902, reported as 89.02% (rounded to 89% in summaries).
• Approach: repeatable transformation and reconciliation steps moved into Power Query and the data model; review and sign-off kept as human steps.

Lessons, constraints and improvement opportunities

• Measuring the baseline properly was what made the result credible; a single “before” timing would have been easy to dispute.
• The largest gains came from removing rework, not from making the visual report faster.
• A natural next step is to monitor input data quality at source, so that exceptions are flagged before the cycle starts rather than during it.