What The Number Does Not Say

A KPI at 43% means nothing. It is a bare fact. An orphaned data point. To speak, it needs a context. And that context is what most steering systems ignore.

The Problem Of The Number Alone

Your dashboard shows an indicator in red. Digital transformation rate: 43%. Target: 60%. Gap: -17 points.

And now what?

You know there is a problem. You do not know why. You do not know what to do. You do not even know whether it is serious.

43% in an industry where the average is 30% is a strong performance. 43% when your direct competitor is at 75% is an emergency. 43% when new regulation requires 80% within six months is a crisis.

The number is the same. Its meaning changes completely.

That is the fundamental problem of contemporary steering. We have data. We lack context.

The Illusion Of Measurement

The last twenty years have produced an explosion of metrics. Companies measure everything. Revenue, margins, NPS, engagement, velocity, turnover, time-to-market, customer satisfaction, conversion rate, cost of acquisition.

And to organize that profusion, a parallel explosion of nomenclatures. KPIs. OKRs. North Star Metrics. Leading indicators. Lagging indicators. Balanced Scorecards. Every era invents its vocabulary, every consultancy its methodology, every guru an acronym.

That semantic inflation gives the illusion of progress. But renaming an indicator makes it neither more accurate nor more actionable. Calling a number a “North Star” illuminates nothing if it floats alone in a void, disconnected from the actions that influence it and from the organization’s real capabilities.

We have made the language of measurement more sophisticated faster than our understanding of it. The problem is not that companies lack indicators — they have too many. It is that they lack links between them. And no acronym, however elegant, will make up for the absence of context.

Dashboards have grown sophisticated. The visualizations are elegant. The data is fresh, sometimes real time.

And yet the leaders we meet share the same feeling: they see a great deal and understand little.

That is not a problem with the quantity of data. It is a problem with the quality of context.

An isolated KPI is like a word pulled out of a sentence. Technically correct. Practically useless. The word “bank” means nothing until you know whether it is the side of a river, a financial institution, or a shoal of fish.

Indicators work the same way. Their meaning emerges from their context.

The Five Dimensions Of Context

Working with organizations across varied sectors — governments, banks, industry — we have identified five contextual dimensions that turn a number into actionable information.

The causal dimension. What influences this indicator? Which other KPIs are linked to it? A high turnover rate in the IT team is not an isolated HR problem. It may be the consequence of project overload. Which is itself the consequence of under-investment. Which is itself tied to budget trade-offs. Without that causal chain, you treat the symptom, not the disease.

The capability dimension. Do you have the skills to improve this indicator? A digital transformation target of 60% is theoretical if your organization has only three senior developers and the market is dry. The KPI does not tell you that. It would have to be connected to your skills map. To your identified gaps. To your real capacity to recruit or to train.

The operational dimension. What are you already doing to improve this indicator? Which actions are under way? How far have they got? Are they sufficient? An indicator in red with three major projects under way does not call for the same reaction as an indicator in red that nobody is addressing. But how many dashboards connect KPIs to the actions aimed at them?

The temporal dimension. Where does this number come from? What is its trajectory? Which past decisions influenced it? An indicator at 43% that stood at 20% a year ago tells a story of progress. The same indicator that stood at 65% tells a story of decline. The instantaneous number hides the dynamic.

The external dimension. What is happening around you? Does new European regulation change the terms? Has a competitor innovated? Has the macroeconomic context shifted? An indicator does not exist in a vacuum. It exists in an environment. Ignoring that environment is steering with blinkers on.

The Leader As Integrator

Faced with that complexity, what do leaders do today?

They integrate by hand.

They look at the KPI on their dashboard. Then they open another system to see the projects under way. Then they call their HR director to understand the state of the skills. Then they ask their strategy team for an update on the competition. Then they try, mentally, to connect all of it.

The leader’s brain becomes the integration system. The human middleware connecting information the systems do not connect.

It is exhausting. It is slow. And it is fragile.

Exhausting because that cognitive integration consumes considerable energy. Leaders describe whole days spent not deciding, but understanding. Gathering the pieces of the puzzle.

Slow because by the time the context has been assembled, the situation has sometimes changed. Information is perishable. A context assembled over two weeks is an obsolete context.

Fragile because the human brain has limits. It forgets. It simplifies. It biases. A leader cannot hold in active memory every connection between every element of their organization.

The Experience Of Integrated Context

For several years we have been exploring a different approach. Not adding more data to the data. But connecting what already exists.

The starting intuition was simple: what if each piece of information carried within itself its connections to the others? What if a KPI “knew” which actions influence it, which skills it requires, which other indicators are linked to it?

Technically that meant rethinking the architecture of the data. Moving from a logic of separate tables to a logic of an interconnected graph. Each node — an indicator, an action, a person, a skill — connected to the others by explicit, traversable links.

The result changes the experience of steering.

A leader clicks on a KPI under strain. They immediately see the three actions under way to improve it. They see that one of them is blocked by a dependency on another project. They see that the project is short of two critical skills. They see that those skills are scarce on the market but that an internal colleague could be trained in three months.

A few seconds. Not weeks.

And above all: they did not have to go looking. The context was there, structurally present, waiting to be consulted.

Artificial Intelligence As A Reader Of Context

Relational architecture solves part of the problem. It connects what is internal to the organization. But external context — regulatory, competitive, macroeconomic — still has to be brought in.

This is where artificial intelligence finds, in our view, its most relevant role.

Not as an oracle that predicts the future. Not as an assistant that answers questions. But as a reader of context. A system able to understand information of different kinds and extract from it what is relevant to a given situation.

New European regulation appears in the official journal. It runs to 200 pages. AI can read it, identify the passages that concern your sector, and flag the indicators on your dashboard that will be affected.

A competitor announces an acquisition. AI can analyse the strategic implications and connect them to your own initiatives under way.

A sector report is published. AI can extract the relevant benchmarks and put your own performance in context.

This is not magic. It is augmented reading. The capacity to process a volume of contextual information no human could absorb, and to surface from it what matters for one specific decision.

From Observation To Action

Context is worth nothing if it does not lead to action.

A recurring frustration among leaders: the gap between understanding and acting. They see the problem. They understand the causes. Then they have to switch to another system to create a task. Call a meeting. Launch a project. The thread breaks.

In our approach, the action is born at the very point where the need is identified. A KPI is under strain? From that view, the leader can create the corrective initiative directly. Assign an owner. Set a deadline. The action is immediately linked to the indicator it targets. The loop is closed.

This is not a question of interface or ergonomics. It is a question of informational integrity. When the action is created elsewhere, in another system, the link to the context is lost. Six months later, nobody knows why that task exists. Which problem it was supposed to solve. Whether it is still relevant.

When the action is born in its context and stays attached to it, it keeps its meaning. It can be assessed against its original objective. It contributes to organizational memory.

The Limits Of Our Experiment

We do not claim to have solved the problem of contextual steering. Our work raises as many questions as it answers.

The first: the risk of information overload. If every KPI is connected to dozens of elements of context, is the leader not going to be swamped? We have learned that there is an art to prioritization. All the context must be available. All the context must not be displayed. The system has to know what is relevant in the moment.

The second: the quality of the links. A context is only useful if the connections are right. A KPI badly connected to its real causes misleads more than it illuminates. Building those links requires domain expertise the system cannot invent. It can suggest. It cannot decide.

The third: human acceptance. Decades of traditional dashboards have shaped habits. Some leaders are disconcerted by an approach that shows them more than they asked for. They wanted a number. They are shown an ecosystem. Adaptation takes time.

These limits are not objections. They are the work in progress.

What Context Changes

We have watched leaders use this approach over several months. A few observations emerge.

The time to understanding falls drastically. What used to take days — gathering the information, calling on the experts, cross-checking the sources — now takes minutes. Not because the information is simpler, but because it is already connected.

The quality of decisions appears to improve. That is harder to measure, but leaders report making decisions they would not otherwise have made. Connections they would not have seen. Implications they would have ignored.

Conversations change in nature. In meetings, less time is spent establishing the facts — the system supplies them — and more time interpreting them. The debate is about what to do, not about what is happening.

And perhaps most important: a form of calm. Several leaders use that word. The sense of no longer steering blind. Of finally seeing the ground they are advancing over.

Conclusion: The Number And Its Shadow

Every indicator casts a shadow. That shadow is its context. The causes that influence it. The capabilities it requires. The actions aimed at it. The environment that conditions it.

The number alone is sharp, precise, reassuring. But it lies by omission.

The shadow is blurred, complex, sometimes unsettling. But it tells the truth.

Steering on numbers alone is steering while ignoring the shadows. It is possible. It is common. It is not enough.

The next frontier of steering is not having more data. We have enough. It is seeing the connections. Understanding the contexts. Grasping at a glance not an isolated number, but the ecosystem that gives it meaning.

That is what we are trying to build. A system where the number is never alone. Where the context is always there, a click away. Where the leader can finally see what their data had been trying to tell them all along.

The number is not enough. It never was.

What counts is what the number does not say.