People Centricity case study

Same KPI Name, Different Clock

53recruiting KPIs on the long list
9regions interviewed on how they measure
8hidden differences in 45 checks
10steps in a method other functions can reuse

The situation

The question behind the question.

Every region of this company reported on hiring. Its new talent acquisition strategy now wanted those numbers in the 2023 scorecards, cascaded from the company to HR to the recruiting team. That raised a question nobody had asked. Each region had built its own dashboards, some on an agency’s definitions, some by hand. A KPI with the same name in 2 countries might not be the same KPI.

As it was asked

“Collect our recruiting KPIs and tell us which to use in 2023.”

As we reframed it

Check that each KPI measures the same thing everywhere before choosing any: then keep the few that pass the tests, and build them into the system so every region reads one clock.

How we worked

How the work unfolded.

01

Collect what is counted

We started with what was already measured. 53 KPIs went into one long list in 8 categories, each with its purpose, formula, source and audience. Regions sent the dashboards they actually used. 9 calls with regional talent acquisition leads, about 14 people, then went through each KPI: how it was defined, how it was calculated, and how much it mattered on a scale of 1 to 10.

02

Find where the clocks differ

Then we compared the clocks. For 5 core KPIs in 9 regions, 45 checks in all, 8 hid a difference behind the same name. One region stopped time to fill when the offer was sent, not accepted; another started it at the intake meeting. One counted working days, not calendar days. Internal hiring was counted only above a certain grade in one region, and included succession moves in another. Only offer acceptance and new-hire attrition held up everywhere.

03

Keep 5, test each

Then the shortlist. Each KPI had to pass 4 tests: it measures an objective of the talent acquisition strategy, someone who can move it owns it, the system can produce it with no manual calculation, and it can be compared inside and outside the company. 5 passed: time to fill, time to hire, internal versus external hires, offer acceptance and new-hire attrition. Each got a one-page card, an external benchmark and its level, from company scorecard to recruiting dashboard. The rule was global definition, local targets: one ruler everywhere, ambition set by each region against its own market.

04

Build it into the system

A definition is only real once the system enforces it. Each KPI was traced to its fields in the HR system, and one gap surfaced: the recruiting process had no offer-acceptance step, so the clock had nowhere to stop. The system partner added it. The dashboard got traffic-light bands and filters by function, location, grade and recruiter. New-hire attrition became the read on whether the new assessment approach hires better. A 10-step method and template let other functions build their KPIs the same way.

The instrument

One card, one clock

The clock. What starts the count, what stops it, and in which unit: calendar days from job posting to offer accepted, for instance, not working days to offer sent.

The source. The system fields that produce it with no manual step, and a published external benchmark to read it against.

The owner. Who answers for it, at which level of scorecard, how often it is reported, and who sets its target.

When 2 countries report the same KPI, they now report the same thing.

What moved

What changed, and what didn't.

5recruiting KPIs live in one system dashboard, defined the same way in every region
Evidence Decision taken Adopted in practice Measured impact
What moved

The global head of talent acquisition validated the 5 definitions in November 2022, and the dashboard went live in the HR system: 5 KPIs, one definition, read by every region monthly and on demand, with no figure rebuilt by hand. The company then set a 3-year plan to work through the backlog of the remaining KPIs with the same method.

What didn't move

Where applications and offers still travel through agencies or email, outside the system, the clock cannot see them. Internal hires were not yet separated from internal moves, and the probation pass rate and diversity KPIs stayed in the backlog. Local targets make the definitions comparable across regions, not the ambition.

What we'd say to a peer

Before you compare recruiting KPIs across countries, check the clocks. The same name often hides a different start, a different stop, or working days instead of calendar days. Fix the definition globally, let each region set its own target, and let the system produce every figure, so nobody rebuilds one by hand.

Let's talk Discover People Centricity Read more cases
Method note:

Own approach in 4 steps; 53-KPI long list; 9 regional interviews; deviation analysis (45 checks); 4 tests (relevance, controllability, technical feasibility, benchmarkability); one-page KPI cards with system logic and external benchmarks; dashboard specification; 10-step reusable method.

WeAreHuman@Work

One piece of evidence on sustainable work, every Sunday. Free, and one click to leave.

Get the weekly edition