How to Start a GEPP Process

In summary

Most GEPP processes fail not because of the method, but because of the start: whatever hasn't been solidified early on weakens everything that follows.

Before any tool, you need a role mapping built through on-the-ground interviews, followed by a skills mapping that goes right down to the observable task itself.

The reference framework should stay heterogeneous site by site at first, before being consolidated: standardising too early means flattening real differences.

The negotiated GEPP agreement (for companies above 300 employees) only has value if the diagnosis it rests on is solid.

Most descriptions of the GEPP process start with the 5 canonical steps: diagnosis, forward-looking analysis, action plan, negotiation, follow-up. That's accurate, but it doesn't answer the question actually facing people in companies tackling this for the first time: in practical terms, where do you start, and what needs to be solid before moving on?

We recently observed this at an industrial company whose technical roles had never had a job description, unlike management positions. These roles were learned "on the job", passed informally from an experienced worker to a younger one, with no document formalising what a team leader or a seasoned operator actually knows how to do. Against a backdrop of turnover and retention pressure, this company started its process from close to zero.

How to start a GEPP process: what it actually means

What to avoid: deciding "we need a GEPP tool" and launching an HRIS project before even knowing what will go into it. A tool doesn't create knowledge of the roles, it organises it, once that knowledge exists.

Step 1: Identify the right people on the ground. Site by site or branch by branch, it isn't always the same profiles who genuinely know a role: sometimes it's a manager, sometimes a senior operator recognised as the reference point. This initial framing determines how reliable everything that follows will be.

Step 2: Map roles and career paths, through interviews. 1 to 2 hours of live conversation with someone who knows their ground is worth a lot. A well-designed questionnaire will help structure the interview. A technical role can't be summed up in a job title: it needs to be told, how someone got there, what they actually do, where they can move on to next. A progression map built this way (for example: assembly assistant → assembler → team leader → site supervisor, with side branches) carries a memory of the role that no generic framework can reconstruct on its own. This logic connects directly with an organisation built around skills rather than fixed job titles.

Step 3: Map skills, area by area. Once roles are set, the next task is to identify, for each one, the real operational skills, organised into areas and then sub-areas (for example, for a technical maintenance role: fundamentals, lockout and protection, routine upkeep, corrective maintenance). Where no job description exists, a common situation for technical roles, an existing generic taxonomy (such as France's ROME directory) can speed up the work, provided it's treated as a starting point to enrich with the role's real specifics, never as a ready-made answer.

One thing changes everything at this stage: how far down do you go? "Proficient in electrical lockout" tells a manager nothing. It doesn't say what to have someone practise, with whom, or how. The real lever is the specific task: the precise, observable action an operator must know how to carry out on the ground. This level of detail is what turns a GEPP process from a one-off project into a living management tool that updates continuously. This granularity shapes everything downstream: gaps identified at the right level, training that's genuinely targeted, individual oversight a manager can use day to day. Critical know-how is often what disappears fastest when this granularity is missing, a subject in its own right that there isn't room to cover here.

Step 4: Complete the career paths: prerequisites and bridges. What remains is to formalise the prerequisites for each transition (a training course, a certification, a qualification), and above all to distinguish 3 quite different types of progression: specialising within the same role, moving up within the same career path, and bridging into a role from a different path. It's this last type that gives the process its real strategic value: it reveals mobility bridges that no one had formalised, sometimes not even identified.

The first realistic milestone isn't a finished framework but a still-uneven map, with terminology that varies from site to site, uneven levels of detail, and areas that remain unclear. Trying to standardise too early means flattening real differences between roles rather than understanding them, a common pitfall across heterogeneous multi-site networks.

Step 5: Consolidate into a single reference framework. Each branch has produced its own version: its own words, its own level of detail, its own repetition. The next step is to bring it all together, align the vocabulary, merge duplicates (two roles named differently but identical in practice), and check that the climb back up from tasks to skills to areas correctly reconstructs the original roles, without gaps or contradictions. This is closer to tailoring than to collecting, but it's what turns scattered pieces of ground-level knowledge into a single, usable structure, the same one for every manager and every training team.

What comes after a well-started GEPP process

Once the role and skills mapping is in place, even imperfectly, 3 movements follow on naturally, each depending directly on the quality of the one before it.

From mapping to gaps. The framework built on the ground only reaches its full value once it's set against the company's future needs: strategic shifts, technological change, recruitment pressure, the environmental transition. This is the forward-looking analysis. If the company's real roles and skills are well mapped, the gap analysis will produce well-calibrated priorities: training goes to the right skills, roles under pressure get addressed, and so on.

From diagnosis to action plan. Training, internal mobility, targeted recruitment: these decisions only carry weight if they're based on gaps that have actually been identified rather than assumed. This is also the stage where the bridges identified in Step 4 take on their full operational meaning: they open up concrete mobility prospects, provided they were formalised early on.

The move to the individual level. At some point, the collective mapping has to translate into individual assessment, a skills matrix per employee, with a real proficiency level. This is a different project from the previous one, more sensitive, and it requires clear HR ownership and firm decisions about how it will be used. It's worth securing the overall structure first before opening this front.

The regulatory outcome, for companies concerned. Above 300 employees, the process ultimately leads to negotiating a GEPP agreement with employee representatives: a three-year agreement, measurable objectives, a monitoring committee. This negotiation can also come with work on funding training pathways. But it only makes sense if everything before it holds up: a GEPP agreement negotiated on a shaky diagnosis remains a compliance exercise, not a real management tool.

What Daylindo brings to this process

Once the role and skills mapping has been built, the difficulty changes shape: it needs to be kept alive, kept up to date, and connected to teams' daily work, rather than left as a document that goes stale the moment it's approved. That's the role the Daylindo platform plays: it structures ground-level skills data down to the specific task, allows assessments to be tracked in a way that holds up as evidence, and gives managers an individual management tool they can use day to day, not just a static document base.

Evidence and feedback from industrial companies that started their GEPP process under these conditions confirm a simple point: how solid the starting framework is directly shapes the quality of everything that follows, from the training plan to the negotiated agreement. Daylindo also supports building professionalisation pathways on this basis, in particular through AFEST (workplace-based training).

The common thread

What sets a GEPP process that delivers results apart from one that stays on paper is almost never how sophisticated the chosen tool is. It's the solidity of what was done at the start, the quality of listening on the ground, the patience not to standardise too fast, and sustained human ownership to turn a map into real pathways for each employee.

FAQ

Where should you start a GEPP process when technical roles have never been formalised?

By identifying the right people on the ground, then mapping roles and skills through direct interviews with those who genuinely know the work.

Do you need a GEPP tool from the outset?

No. A tool organises knowledge of the roles that has to exist first. Launching it before the ground-level mapping means digitising an empty shell.

How far down should the skills mapping go?

Down to the specific task: the precise, observable action an operator must know how to carry out on the ground. This is the level of detail that makes a GEPP process manageable day to day, not just descriptive.

Can you rely on an existing taxonomy such as ROME?

Yes, provided it's treated as a starting point to enrich with the role's real specifics, never as a ready-made answer.

Are all companies subject to the negotiated GEPP agreement?

The obligation to negotiate a three-year agreement with employee representatives applies to companies with more than 300 employees. Below that threshold, the process remains relevant but without this regulatory requirement.

How long does it take to lay the groundwork for a GEPP process?

There's no standard timeframe: it depends on the number of roles and sites to map. What matters most is usually not speed, but the quality of the ground-level interviews and the patience not to standardise a still-uneven map too soon.