Employee upskilling: a plan and metrics for growing companies

Employee upskilling is the structured process through which a company closes the gap between its people's current skills and the skills that the evolution of their work requires. It is not a series of courses, but a plan that links gap assessment, a learning path, protected application, and behavioral measurement.

In the ISTAT (Italian National Institute of Statistics) survey on training in companies, one third of Italian companies with at least 10 employees report that some of their staff lack the skills required for their work, and they point mainly to technical and operational skills (32.0%), teamwork (31.2%) and problem solving (29.8%) as the gaps to close [1]. The figure marks out a priority area of work.

The next sections cover the definition, the phases of the plan, effectiveness metrics, and common mistakes.

Understanding what an upskilling plan in a company really is

The term "upskilling" is used for very different things: from a single course to a structured multi-year plan. The context, however, is clear: in Italy's Excelsior survey, planned hires that are hard to fill rose from 26.4% in 2019 to 47.8% in 2024, and the strategy companies choose most often to deal with the problem — in roughly one case out of two — is to look for people with skills similar to those required and, after they join, put them through targeted training paths [2]. Reducing the ambiguity is the first step: a useful upskilling plan is a structured process, not a collection of scattered training activities.

Before going further, it is worth clarifying three recurring distinctions. Upskilling develops new skills within the same role: an administrative employee who learns to use advanced digital tools stays in the same role, but with broader capabilities. Reskilling changes the team member's role, for example a production worker who becomes a quality inspector. The two responses have different costs, timelines and risks: confusing them leads to poorly calibrated paths. Second distinction: continuing education is the recurring update of skills (mandatory courses, refreshers, regulations) — ordinary and structural. Upskilling, by contrast, is an intervention aimed at closing a specific gap identified upfront: strategic, not routine. Third distinction: onboarding works with new team members in their first 30-90 days; upskilling works with team members who are already in place, to bring them to a higher level of skill in the role they already hold. Someone in an upskilling path is not "new."

A series of annual catalog courses is not an upskilling plan: without a gap assessment upfront and behavioral measurement afterward, training activities do not develop skills in the role, only generic updates. The distinction is not academic — it has direct consequences on the return on training investment. Professional growth is one of the strongest motivational levers in an organization; to explore the link between development and motivation, read how employee motivation works.

Assessing the skill gap before choosing the path

Skipping the gap assessment is the most common cause of upskilling paths that miss the target. Without a need defined upfront with explicit criteria, the choice of path ends up relying on the available catalog rather than on the actual work, and the result becomes unverifiable from the start. The assessment relies on three tools: analysis of the actual work (what needs to be done), evaluation of current skills, and identification of the critical gap (not just the desirable one).

The first tool is observing the actual work. You start from an operational question: which tasks require skills that the team member does not have, or has to an insufficient degree? You do not start from a list of abstract skills taken from a catalog, but from the concrete work the team member does or will have to do. A useful reference document at this stage is the role description: if the company has a structured job description, the gap analysis becomes much more precise (see how to structure a job description).

The second tool is evaluating current skills. In a company of 10-15 people, this does not mean formal assessment centers, but a structured conversation with the team member, together with direct observation by their manager. The goal is to map the current level on each skill relevant to the role — not an overall performance evaluation, but a specific snapshot of the target skill.

The third tool is identifying the critical gap. Not all gaps have the same urgency or impact. A critical gap is one that, if not closed, causes repeated errors, operational slowdowns or lower-quality decisions. A "desirable" gap is one that would improve performance but does not compromise it if left open. Building the plan on real urgency, not on ambition, is the principle that separates a functional upskilling plan from one that is educational but wishful.

Even in a company with 15 employees, one hour of structured observation is far better than a "what do you need?" survey sent by email. The subjectivity of team members' answers about their own gaps is well documented: people tend to underestimate the skills they lack in their own domain and to overestimate those in the areas they see as their favorites.

Building an upskilling plan in 5 workable phases

A useful upskilling plan, stripped to the essentials, follows a five-phase sequence: gap assessment, choice of training method (internal/external, individual/group, synchronous/asynchronous), design of protected application, execution of the path, and behavioral measurement at 3-6 months. Skipping the fifth phase is the most common route to plans that "seem to work" and then produce no change. The order matters: the assessment always comes before the choice of method.

Phase 1 — Gap assessment. As described in the previous section: observation of the actual work, evaluation of current skills, identification of the critical gap. Indicative duration: 1-2 weeks for a single role.

Phase 2 — Choice of training method. The method is chosen based on the type of skill to develop, not on the available catalog or on the cost of the path. For specific technical skills (software, procedures), the combination of an external course plus internal application is often effective. For soft skills (communication, handling feedback, coordination), structured coaching on the job outperforms classroom days.

Phase 3 — Protected application. Transferring learning to real work does not happen automatically. Protected application is a space where the team member can practice the newly acquired skill in real but low-risk situations: real tasks, with active supervision and structured feedback. Without this phase, most training content does not survive beyond three weeks of active memory.

Phase 4 — Execution of the path. The execution of the plan — training sessions, modules, on-the-job coaching — follows the designed sequence. Duration depends on the complexity of the skill: paths on narrowly defined technical skills are completed in 4-8 weeks; paths on complex soft skills take 6-12 months. Discontinuity (stopping and restarting the path) is one of the factors that wipes out the training result.

Phase 5 — Behavioral measurement at 3-6 months. Not course satisfaction. Not the final test. But checking, 3-6 months after the intervention, whether and how the team member applies the acquired skill in real work. This phase closes the plan. Management training for people who coordinate others is an area where the 5-phase sequence produces particularly measurable results; to learn more, read how management training works.

Choosing the right mix of training methods for each gap

Different skill gaps require different methods. For specific technical skills, the combination of an external course plus internal application often works; for soft skills, structured on-the-job coaching with structured feedback is more effective; for digital skills, distributed microlearning can outperform isolated intensive days. Choosing the mix requires reading the nature of the skill before the course catalog.

A guiding matrix can help make this choice systematically:

scroll the table →

The principle behind the choice is this: the skill determines the method, not the other way around. A one-day course on a complex skill can hardly be classified as effective training: without protected practice between sessions, the content stays at the level of declarative knowledge, not applied skill. Alternating content, application and feedback is not a methodological refinement, either: in 2020 about half of Italian companies with at least 10 employees used training activities other than courses — training in the work situation, coaching and job rotation, attendance at conferences and seminars — an increase of about 10 percentage points compared with 2015 [1]. To explore the mechanisms of adult learning, read how effective adult learning works.

A note on microlearning: the format is effective for basic digital skills and procedural updates, but it is an operational hypothesis not yet backed by studies that systematically compare it with intensive days in small Italian companies. Treat it as a direction to test in your own context, not as a universal principle. That this row of the table matters, however, is documented: in Italy fewer than half of people (45.7%) have at least basic digital skills, against a European average of 55.6%, and Italian companies with fewer than 10 employees are the size class reporting the greatest difficulty in finding people with mathematical and IT skills, with 57.1% of hires for which that skill is considered a priority [3].

Measuring upskilling effectiveness beyond the number of training hours

Measuring upskilling only by the number of hours delivered or by participant satisfaction is late and misleading. In practice, in small and mid-sized Italian companies, training is measured more often at the first level — participant reaction — than through checks on behavior and results. You need indicators that close the loop: observable changes in work at 3-6 months, the quality of decisions in which the learned skill comes into play, and fewer errors in real work.

Kirkpatrick's conceptual model (1959) identifies four levels of measurement: reaction (satisfaction), learning (declarative acquisition), behavior (transfer to real work), and results (impact on the organization). The vast majority of smaller companies measure only the first level because it is the cheapest to collect. Levels 3 and 4 require observation months later, but they are the only ones that produce information useful for deciding whether the plan worked.

A plan measured on a single metric is blind; one with ten indicators becomes opaque. Three balanced indicators are a readable set for most companies. Here are three examples by functional area:

Sales: fewer errors in the quotes sent; a higher conversion rate in negotiations where the developed skill was applied; perceived quality of written feedback to customers (sampled at 6 months).

Operations: fewer rejects or procedural errors in the weeks after the plan; execution speed on processes where the developed technical skill is directly involved; the ability to independently train new team members on the same process (a sign of consolidation).

People/coordination: frequency and quality of the structured feedback the team member gives to their direct reports (direct observation by their manager); fewer escalations on manageable operational conflicts; improved clarity in written delegation.

The number of training hours delivered is an administrative metric, not an effectiveness metric. Recording them is necessary for regulatory requirements and for training-fund reporting, but it says nothing about actual skill development.

Common mistakes in company upskilling plans

The most common mistakes in upskilling plans are not technical, but process-related: choosing the method before assessing the gap, compressing paths into intensive days without protected practice, evaluating only on participant satisfaction, applying methods designed for large companies in contexts where they do not work (because of group size, duration, mixed levels), and abandoning the plan after the first months without behavioral measurement. Recognizing them is more useful than memorizing them, because they show up in different forms depending on the size of the company.

Mistake 1 — Choosing the method before assessing the gap. This is the most common mistake. Someone decides that "we need a communication course" without having identified which specific communication gap is causing friction. The operational fix: no training path starts before the critical gap has been defined in writing.

Mistake 2 — Concentrating the path in intensive days without protected practice. A two-day intensive course on a complex skill has a very high decay rate without structured application in the following weeks. The fix: always design a unit of protected practice for each training module delivered.

Mistake 3 — Evaluating the plan only on participant satisfaction. The feedback "I liked the course" measures neither learning nor transfer. The fix: add at least one behavioral indicator collected 3 months after the intervention.

Mistake 4 — Applying methods designed for large organizations in small companies. Paths designed for homogeneous groups of 20-30 participants do not work with 5 people at different levels. The fix: scale the method to the actual structure, even at the cost of simplifying it.

Mistake 5 — Not identifying the "critical" gap among the available gaps. When several gaps emerge at the same time, the tendency is to try to cover them all. An effective plan works on the gap that, if not closed, causes the greatest operational damage. The fix: rank the gaps by urgency and impact before building the plan.

Mistake 6 — Abandoning the plan after the first months without measurement. The plan stops under operational pressure, and nobody checks whether the skill was acquired. The fix: schedule the behavioral measurement session at 3-6 months before the path even starts, and treat it as a binding commitment.

Limits and conditions of applicability

The guidance in this article applies most precisely to companies with a recognizable organizational structure (at least one coordination level and relatively well-defined roles). In micro-businesses of 2-3 people, the distinction between upskilling, continuing education and informal learning tends to blur, and the operational phases of the plan can be oversized for the context.

The 5-phase model described in the third section is an operational guideline, not a universal protocol. The duration and intensity of each phase vary significantly depending on the complexity of the skill, the team member's available time, and the company's internal resources. Highly operational contexts (production, logistics) require specific adaptations compared with more cognitive contexts (sales, administration).

The institutional sources cited (ISTAT, Unioncamere-Excelsior) provide aggregate national data for Italy, and the ISTAT survey on training covers only companies with at least 10 employees; Italian companies vary widely by sector, size and geographic location. The data do not transfer automatically to an individual company.

Frequently asked questions

Does upskilling also apply to team members with many years of experience? It does, but with some differences in approach. An experienced team member tends to resist paths they perceive as basic, and responds better to methods that use their experience as a starting point (mentoring, on-the-job coaching, real cases) than to standard lecture-style courses.

Is there public funding to finance upskilling plans? In many countries, sector or joint training funds finance company training plans, including skill development paths. Access usually requires a structured training plan and reporting of the hours delivered. For up-to-date information, check with the training fund or body that covers your sector.

How long does a complete upskilling plan take? It depends on the complexity of the target skill. For narrowly defined technical skills (for example a specific piece of software), a cycle can be completed in 4-8 weeks. For complex soft skills (managerial communication, handling feedback), the cycle requires 6-12 months of structured work with continuous application.

Is behavioral measurement at 3-6 months necessary in every case? For technical skills that can be verified directly on work output (product quality, procedural speed), measurement can happen earlier and more directly. For interpersonal and decision-making skills, a check at 3-6 months is the only time frame that lets you observe the change in natural, not simulated, conditions.

Can you build an upskilling plan without an in-house HR manager? Yes, it is workable with a dedicated line manager. The condition is that this person has enough time for the gap assessment, for overseeing protected application, and for follow-up measurement. Failing to oversee these three phases is more critical than the absence of a formal HR function.

Operational summary

An upskilling plan is built in five phases: assessment of the critical gap, choice of method based on the type of skill, design of protected application, execution of the path, and behavioral measurement at 3-6 months. Each phase is necessary; skipping the fifth — measurement — makes the plan unverifiable.

The main operational points:

• Distinguish upskilling (same role, broader skill), reskilling (change of role) and continuing education (ordinary update): the distinction drives very different choices in cost and time.
• Assess the gap with three tools: observation of the actual work, evaluation of current skills, identification of the critical gap (not just the desirable one).
• Choose the method after the assessment, not before: external course plus application for technical skills; structured on-the-job coaching for interpersonal skills; distributed microlearning (as a hypothesis to test) for basic digital skills.
• Measure with three balanced indicators at 3-6 months, not with hours delivered or participant satisfaction.
• Avoid the six most common mistakes: method before gap, concentration without practice, evaluation based only on satisfaction, methods not scaled to a small organization, failure to prioritize the critical gap, abandonment without measurement.

Conclusion

A useful upskilling plan is not a series of courses, but a structured process that links gap assessment, a deliberate choice of method, protected application, and behavioral measurement at 3-6 months. Building it requires letting go of the belief that "we did some training" is the same as "we built skills," choosing the method after assessing the gap, and accepting behavioral measurement as proof of change.

The thread that ties these steps together is the consistency between the company's real need and the path built for the team member. When that thread breaks, upskilling becomes a training cost with no return, and the skill gap goes back to creating operational friction. To place upskilling within the broader training framework, read the guide to company training and, on the principles of adult learning, how effective adult learning works.

A company that genuinely works on upskilling stops being at the mercy of talent shortages on the job market. Existing team members acquire new skills within their role, operational decisions improve in quality in the months after the plan, and replacement costs go down. It is a calmer way of working and a more solid path to growth — accessible to organizations of any size, as long as the plan stays connected to the actual work.

Sources and references

[1] ISTAT, "Formazione nelle imprese — Anno 2020", Istituto Nazionale di Statistica, December 30, 2022. Available at: https://www.istat.it/it/files//2022/12/REPORT-formazione-imprese.pdf

[2] Unioncamere — Ministero del Lavoro e delle Politiche Sociali, Sistema Informativo Excelsior, "La domanda di professioni e di formazione delle imprese italiane nel 2024", Unioncamere, 2025. Available at: https://excelsior.unioncamere.net/sites/default/files/pubblicazioni/2024/Domanda_professioni_formazione_imprese.pdf

[3] Unioncamere — Ministero del Lavoro e delle Politiche Sociali, Sistema Informativo Excelsior, "Le competenze digitali — Analisi della domanda di competenze digitali nelle imprese, indagine 2024", Unioncamere, 2024. Available at: https://excelsior.unioncamere.net/sites/default/files/pubblicazioni/2024/Competenze_Digitali_2024.pdf