GEPP Dashboard: The HR Metrics That Really Matter

In summary

Most GEPP dashboards track the resources committed well (budget, hours, participation), but shed little light on what actually matters: real learning and how fast the organisation adapts.

Managing GEPP covers 4 different levels: resources, skills, learning, and adaptation. The first 2 are usually well measured, the last 2 almost never.

Traditional GEPP manages a stock (a snapshot of skills at a given moment), whereas real skills are built as a flow, in day-to-day work.

Before adding more indicators, you often need to build a skills "factory" rooted in real work: without it, there's nothing to measure.

"Show me your metrics, and I'll tell you what you're actually managing."

Gerardo Marcotti, Founder of Daylindo (www.daylindo.com)

Ludovic Taphanel, Secretary General of ADevComp (Association for Skills Support and Development)

During a workshop on managing GEPP with around 15 HR directors and training managers, we started with a deliberately simple question: what are the 5 metrics you actually look at to manage your GEPP?

The answers came quickly: training budget, hours delivered, participation rate, certifications, age pyramid, turnover. These are the available metrics, the ones people know how to produce, compare, and present to the leadership committee.

Then another question came up: what do these metrics actually tell us about the organisation's ability to adapt?

We're fairly good at tracking what's been organised. We're much less able to say what's been learned, what's changed in practice, which skills are emerging on the ground, or how fast a team is genuinely becoming autonomous.

It's not that the classic metrics are useless. They remain necessary. But they mostly shed light on the resources committed. To manage a GEPP process in a context where roles are evolving fast, you also need to look at something else: real learning, mastered tasks, peer-to-peer transmission, adaptation speed.

The question isn't just: which metrics should we add to our dashboard? It's rather: what are we actually trying to manage?

What our metrics reveal, and what they leave in the dark

2 columns on a flipchart

When participants' metrics were pooled together, 2 columns emerged on their own.

On one side, what organisations know how to measure: budget, participation, hours, certifications, macro-skills, age pyramid. Available, comparable, auditable data.

On the other, what these metrics leave in the dark: activities that are evolving, informal learning, new practices emerging on the ground, skills in the process of forming, weak signals.

If 80% of our metrics manage resources, why be surprised that managing adaptation is hard?

4 questions, not one

What we're actually managing when we manage GEPP

There's a reason for this imbalance. GEPP doesn't cover a single level of management. It covers 4, and they're not measured the same way.

Are we doing what was planned? → Managing resources. Budget, hours, participation rate, number of pathways. The easiest level to measure, and by far the most documented.

Do we have the skills we need? → Managing skills. Frameworks, mappings, manager assessments. Harder to produce. Often self-reported. Quality depends heavily on who's assessing and against what scale.

Are employees really learning? → Managing learning. What actually happens on the job: AFEST, mentoring, observable tasks. Rarely tracked. Rarely recognised. Yet this is where skills are actually built.

Is the organisation learning fast enough? → Managing adaptation. Time to competence, coverage of critical skills, transformation speed. The most strategic level, and the least measured.

Most companies manage level 1 precisely. Level 2, with rough approximations. Levels 3 and 4, almost not at all. This isn't a tooling problem. It's a question of focus, and often, of data that simply doesn't exist yet.

Stock or flow?

The real nature of the problem

There's a deeper tension behind these 4 levels.

Traditional GEPP manages a stock: you map what you have, compare it to what you'd want, and plan training to close the gap. It's a snapshot logic.

Except real skills aren't built that way. They're built through work, through practice, through feedback received on the ground, through transmission between colleagues, through repetition in real situations. It's a flow logic.

Managing a flow means watching the cycle as it happens:

• a real work situation creates an opportunity to learn

• practice produces an experience

• feedback turns it into learning

• progress becomes visible

• recognition consolidates it

• transmission spreads it

If you want to know whether skills are progressing in your organisation, this is the cycle to watch, not the training hours consumed.

An often underestimated problem

Managers and training teams, 2 languages in the same organisation

Before even talking about metrics, there's a more basic problem to fix in many organisations.

Managers and training teams don't always speak the same language when they talk about skills.

The manager thinks in macro terms: can this person hold this role? They want a simple reading: acquired or in progress. That makes sense: their job is to assign people to roles, not to write assessment grids.

The training team, on the other hand, makes sure that transmission of job-specific tasks actually happens. They work on granular, observable micro-skills, measurable in real situations, that evolve with the role. This is the level at which learning actually happens.

Both levels are legitimate in their own way. The problem is when they coexist without being connected. The "micro-skills" acquired and proven in training generally don't feed into the macro view the manager signs off on. The HR mapping doesn't go down to the observable task. And each side's metrics tell the other nothing.

Building a real GEPP management system starts with solving this language problem, and the interaction between operations and training.

What this looks like in practice

At a major telecoms operator: starting with the factory

In 2019, the internal technical skills academy of a major telecoms operator set itself a simple goal: let every employee acquire the skills needed for their current or future role. The intention was clear. The reality, a bit more complicated.

At the outset, 4 findings stood out:

• We measure resources, not results. Training hours, completions, hot-take evaluations, but no data on skills actually acquired.

• Informal learning exists, but it's invisible. The 70/20/10 logic is well known. But what happens in the 70 is neither tracked, nor recognised, nor valued.

• Role frameworks exist, but they live in files, not in real work.

• Roles are evolving faster than the frameworks, with a growing share of skills that don't exist yet today.

The decision made wasn't to build a better dashboard. It was to first build a skills factory: pathways rooted in real work, internal transmitters trained to observe and assess, an approach solid enough to be scaled.

"Before the metrics, you need a factory. You can't measure what you're not producing yet."

The start was modest: fewer than 20 people, a proof of concept from 2020 centred on tooling AFEST, using this format as a Trojan horse for the wider approach. Not a big project. A willingness to change logic, bringing training closer to the ground.

5 years on, the approach has evolved into a systemic logic: skills-based learning, a learning organisation built around skills, management driven by skills flows.

The metrics that changed

When you manage a flow, you measure something else

Once the factory was in place, the metrics changed in nature. They no longer measured what had been organised. They measured what was actually happening. 4 families emerged from this experience.

Rollout: is the approach moving forward? 

Not yet a measure of skills. A measure of the organisation's real engagement. Number of active business units/sites, number of pathways launched/completed, top/bottom entities by activity level.

Skills: what can we actually do? 

Not what's written in the framework. What's actually assessed, proven, dated. Number of skills in the framework, number of skills actually assessed, number of employees actually assessed, coverage rate of critical skills.

Effectiveness: are we progressing, and how fast? 

The hardest to produce. And the most useful for decisions. Average progression gap, time to competence, how levels change over time.

Learning organisation: does the organisation learn on its own? 

The most strategic level. It measures whether transmission holds up without central intervention. Number of trained/active transmitters, transmitter activity rate, spread of practices between peers.

The difference with classic metrics: these 4 families measure what's happening in real work. They assume a ground-level factory exists. You can't produce them with an LMS alone, or with a framework that lives in a file.

What Daylindo brings to this management approach

Building these 4 families of metrics requires reliable ground-level skills data, tracked down to the specific task, not just a framework living in a file. That's precisely what the Daylindo platform structures: it keeps a record of each assessment that holds up as evidence, connects the manager's macro view to the granularity of the task observed by ground-level transmitters, and feeds all 4 management levels rather than just the resources level.

Evidence from industrial companies that have built this kind of factory confirms the workshop's finding: it's flow data, generated through real work as it happens, that makes a GEPP dashboard genuinely useful for deciding, not just for reporting. This approach connects directly with the steps described in our article on how to start a GEPP process.

To conclude

Start with the questions

What this workshop highlighted isn't that HR directors lack metrics. It's that they often lack metrics that answer the right questions.

The useful questions aren't "what's our training rate?" or "how many hours have we delivered?". They are: is our organisation learning fast enough to adapt? Are employees really acquiring the skills the roles need? And how would we actually know?

The answer doesn't start with a dashboard. It starts with a decision: bring training closer to the ground, root pathways in real situations, train transmitters capable of observing and assessing. And progressively build the data that's missing.

You don't need to build everything at once. You can start small, with one scope, one business line, one specific task. What matters is starting with the right questions.

FAQ

Why aren't the classic metrics on a GEPP dashboard enough?

Because they mostly shed light on the resources committed (budget, hours, participation), not on real learning, mastered tasks, or the organisation's adaptation speed.

What are the 4 management levels of a GEPP process?

Managing resources, managing skills, managing learning, and managing adaptation. The first 2 are usually well measured, the last 2 much less so.

What's the difference between managing a stock and managing a flow of skills?Managing a stock means mapping skills at a given moment, like a snapshot. Managing a flow means tracking the real learning cycle: practice, feedback, recognition, peer-to-peer transmission.

Should you build a dashboard before or after the skills factory?

After, or at least in parallel. Without a skills factory rooted in real work, there's no reliable data to measure, even with the best dashboard.

Why do managers and training teams often speak a different language about skills?The manager thinks in macro terms (acquired or in progress), while the training team works on granular, observable micro-skills. The problem arises when these 2 levels aren't connected.

Which metrics should you track once a skills factory is in place?

4 families: rollout of the approach, skills actually assessed, effectiveness of progression (time to competence), and the organisation's ability to pass on skills without central intervention.