Short answer
A useful people-data request tells the analyst which decision it supports, which people and period are in scope, how each measure is defined and who will use the result. Agree the minimum fields and quality checks before extracting data. Keep missing values, corrections and differences between populations visible so the resulting briefing can be interpreted correctly.
Request card: send these fields together
Ask for what the decision needs, with an agreed audience and delivery format. Let the source owner identify restrictions or a more suitable existing measure before creating a new extract.
- Question: the decision, meeting and responsible owner.
- Population: included roles, locations and statuses, plus explicit exclusions.
- Period: snapshot date or start and end dates, with relevant time-zone rules.
- Measures: requested fields and why each is needed.
- Audience: who receives the output and at what level of detail.
- Delivery: format, deadline and secure location agreed with the owner.
- Review: contact for definitions, missing data and corrections.
Definition sheet: one entry per measure
Agree these fields with the person who knows how the source is maintained. Keep the definition beside the output so a later reader does not have to reconstruct it.
- Name and plain-language meaning.
- Unit: people, hours, events, rate or another defined quantity.
- Calculation: numerator, denominator and treatment of duplicates where relevant.
- Source and owner: authoritative record and extraction date.
- Inclusions and exclusions: statuses, locations and events counted.
- Missing values: how unknown, not applicable and zero differ.
- Change note: any definition or source change that affects comparison.
Acceptance checks before using the extract
Check that the returned population, dates and units match the request. Reconcile totals with the source owner where appropriate. Inspect missing records and duplicates using the relevant business key, and confirm how corrections will be supplied. A passed check means the extract meets that check, not that every later interpretation is justified.
Record a mismatch rather than repairing it silently
Write the field, observed issue, possible effect and owner. If two teams use different definitions, keep both names visible until they agree which serves the question. Do not turn an unknown into zero or average percentages without examining the denominators.
Hand the result into the briefing
Attach the definition sheet and quality note to the permitted output. The HRBP can then combine it with employee accounts and operational context. Reuse the request for a later period only after checking whether roles, sources or definitions changed.
Why this guide is useful
Use this original request template when an HRBP needs information from HR operations, finance, an analyst or a process owner. It produces a clear handoff and a small definition sheet. It does not replace the later briefing: the linked HRBP data resource explains how to combine the returned information with other evidence.
Agree fitness for the question
The UK Government Data Quality Framework treats quality as fitness for purpose and asks teams to communicate limitations across the data lifecycle. This template applies that principle to an HRBP request. A clean-looking export may still use the wrong period, population or definition for the decision.
Concrete example
Fictional example: an HRBP asks for “headcount by store” to discuss supervisor coverage. Finance returns 72 full-time equivalents, while HR operations returns 90 people. Neither number is necessarily wrong. The request is revised to distinguish people, contracted hours and coverage periods. The meeting can then discuss which measure actually informs supervisor workload, without silently substituting one count for another.
Frequently Asked Questions
- Which people data should an HRBP request first?
- Start with the decision and select the minimum information that can inform it. Agree the population, dates, measures and audience with the source owner before asking for a broad export.
- How should missing people data be handled?
- Distinguish unknown, not applicable and zero. Report the missing records relevant to the question, ask the owner about their cause and state the limit on any comparison or calculation.