People Strategy case study

Many HR Reports Already Existed. Nobody Knew.

~300group HR metrics in the baseline, across 8 topics
5 + 5key markets and houses mapped against the group set
24hthe reload cycle that decided which tool does what
4operating models compared for people analytics

The situation

The question behind the question.

This luxury group was moving its HR data to a new core system, with an analytics tool beside it and a go-live set for the last quarter of 2024. The tools were meant to show whether its people strategy was delivering value. Yet reporting had grown the way the group had: at headquarters, then market by market and house by house, some 300 metrics at group level alone. The systems had a date. The decisions behind them did not: which numbers matter, where each is answered, and who runs analytics afterwards.

As it was asked

“Help us deploy our new HR reporting and analytics tools.”

As we reframed it

Treat the tools as the enabler, not the goal: decide which numbers show the people strategy delivering value, from the centre out, where each is answered and for whom, and who runs analytics after go-live.

How we worked

How the work unfolded.

01

Map from the centre out

We opened with a 1-month diagnosis: the business and its operating model, its HR technology today and planned, and the other workstreams of the HR system programme. Then the baseline, in 3 rings. First, every HR metric reported at headquarters, for business and compliance reasons: some 300, across 8 topics. Then 5 key markets, then 5 houses, each mapped against that group set, with every local specific sorted as compliance-driven or not. Usage added a cut: 6 group dashboards carried 82% of all use; more than 30 local ones were rarely opened.

02

Place each question

Then the split between the 2 tools, Workday and Visier. The rule was freshness. A question that needs live information on the workforce goes to the core HR system, where employees, managers and HR find their reports. The analytics tool reloads every 24 hours, so it takes trends and visuals, for HR. Every request was ranked the same way: first its link to a people strategy objective and its owner’s reach, then feasibility and effort, as assessed by the implementation partner. The HR leadership team made the trade-offs. We worked with that partner and both vendors on what each tool could deliver and how to configure it.

03

Design the model behind it

A tool does not run itself. We compared 4 operating models for people analytics against 5 questions, from scope to maturity and budget, and the group chose a lean central team with expertise placed in the markets. It mirrors how HR already supports the business. Around it we set how reporting requests and issues are escalated, a data security matrix of who sees what, and 2 change strategies, each with its own communications and training: one for HR, which would work differently, and one for the business, which mostly needed to find its live reports.

04

Test before go-live

We project-managed the implementation partner from requirements through configuration to user acceptance. There, every indicator and dashboard was checked against the baseline, every discrepancy traced to its root cause and resolved, and testers signed in under different roles to confirm each saw only what the security matrix allowed. We also set how adoption would be judged, from data integrity to weekly active use, with corrective actions agreed in advance for any market or business that lagged. In parallel, one of the group’s houses asked us to go deeper on the dashboard for its managers.

The instrument

3 rings, 1 escalation route

The centre. A lean team owns the group set of metrics, the 2 platforms, the data security matrix and the priorities. It stays small on purpose.

The markets. Analytics expertise sits in each key market, close to the business, working to the centre’s standards.

The houses. Each house is served by its market. A local metric stays when compliance requires it; any other is weighed against the group set before it is built.

A request or an issue travels the same rings in reverse: from house to market, and from market to centre, where the group set and the platforms are kept.

What moved

What changed, and what didn't.

10work streams managed under one plan, from the first baseline to user acceptance
Evidence Decision taken Adopted in practice Measured impact
What moved

We managed all 10 work streams under one plan, closest on strategy and quality, lighter on the technical build. The group went live in the last quarter of 2024 with both tools and its chosen operating model: a lean central team, expertise in the markets. A new global head of people analytics took over after 3 months of shadowing.

What didn't move

The markets and houses took a month longer than planned to send and rank their reporting needs, pushing design and testing back a month. The go-live date held; the buffer did not. More than 30 local dashboards were retired, not rebuilt: most answered one-off requests already covered by dashboards their users did not know existed.

What we'd say to a peer

A new HR system is an enabler, not the result. Start from the numbers that show the people strategy delivering value, then work from the centre out, letting a local metric in only where the law requires it or the group set lacks it. Put live and trend questions in different places, check whether an answer already exists before rebuilding a report, and name who runs analytics before the project leaves.

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

Own approach, 4 steps; 3-ring metric baseline; 4 operating models tested on 5 questions; indicator checks at user acceptance.

WeAreHuman@Work

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

Get the weekly edition