Govern KPIs
The framework that makes business KPIs stable, shared, and consistent over time.
When KPIs multiply without a system governing them, the problem is not having too little data — it is that every team has too much of it, everyone interprets it in their own way, and strategic decisions are made without a common language. We build the framework that unifies performance measurement at company level: from North Star Metrics for the board to diagnostic trees for operational teams, through to the semantic model that makes those KPIs stable, shared and non-negotiable over time.
In mature organisations, the problem is not having too few KPIs — it is that every team has too many, calculates them in its own way and nobody formally owns them. Strategic decisions are made without a common language, and the data that should align the organisation instead becomes a source of conflict. We build the framework that resolves this problem at the root: from North Star Metrics for the board to diagnostic trees for operational teams, through to the semantic model that makes those KPIs stable, shared and non-negotiable over time.
What we build together
We start from what exists — reports, dashboards, tools, metrics in use — and build a real map of the current state. The objective is not to add KPIs but to reduce them to the ones that truly matter, eliminating redundancies and resolving interpretive conflicts between teams.
We define the first-level metrics to bring to the board — balanced between lagging and leading indicators — and build the hierarchical structure that connects them to the levels below: company level, team level, squad level. Every metric has a unique definition, a formula, an owner and an executor.
When a North Star Metric does not move in the right direction, the team needs to know where to look. We build the diagnostic trees that decompose each metric by mathematical and causal factors, making it immediate to identify drivers and responsibilities.
The framework produced becomes the foundation of the semantic layer: a fixed, shared KPI dictionary managed through structured tools. From that point on, data speaks the same language across the entire organisation — regardless of who is looking at it and with which tool.
We design the hierarchy of information assets: which dashboards to maintain, which to deprecate, which to modify. The objective is that every level of the organisation has access to the right information, in the right form, without overlaps or conflicts.
How we do it
1. Discovery
We collect the KPIs in use, analyse existing reports and strategic documents, and run sessions with all relevant teams. The starting point is not a hypothesis — it is a precise picture of how the organisation measures performance today.
2. Rationalisation and framework design
We define the North Star Metrics, build the hierarchical KPI structure and resolve interpretive conflicts between teams. Every metric is documented with definition, formula, dimensions of analysis and ownership.
3. Diagnostic tree building
We develop diagnostic trees for the main metrics, integrating both mathematical decomposition and causal logic. This is the level that transforms a KPI from a number into a governance tool.
4. Semantic model and dashboard governance
We build the semantic layer on the basis of the validated framework and define the hierarchy of information assets. Existing dashboards are mapped against the new system and subjected to feasibility analysis.
5. Handoff and internal adoption
The framework is delivered with all the documentation needed for the team to maintain it, update it and defend it internally. Governance does not depend on our continuous presence.
Answers to your questions
Everything you need to know about how we work and how we can help.
Because this work sits between the technical and the business, and requires someone who does not belong to any specific team. When KPIs are the subject of internal interpretive conflicts, you need someone who has no stake in the game. It is the reason why even the most digitally advanced organisations choose an external partner for this phase.
Exactly there. Discovery starts from what exists, not from a blank slate. Accumulated complexity is the working material, not an obstacle.
Not if it is built correctly. The semantic model is designed to be stable — KPIs that enter the dictionary are fixed and managed through structured tools. We provide the criteria for updating it when the business evolves, but the load-bearing structure does not change with every planning cycle.
It depends on the size of the organisation and the number of teams involved. A first complete cycle, from discovery to delivery of the semantic model, is generally between three and five months.
For structured organisations with at least 3–4 teams that produce reports and dashboards independently, with recurring conflicts over KPI definitions, formulas or interpretations. Typically with a board or C-level that requests aggregate metrics and operational teams that struggle to provide them consistently. It works for those who want to build a common language around data — not just new dashboards.
When the organisation is small or has a single data team that already manages everything in a coordinated way. It is also not the right choice for those expecting to get “the right KPIs” without a preliminary discovery process on real workflows, for those looking for a new BI tool or a new dashboard (that is an output operation, not a governance one), or for those seeking a cosmetic rationalisation exercise without the mandate to resolve the interpretive conflicts that already exist.
Let’s talk about your
next project.
Thirty minutes to understand where to start and what makes sense to build.