Employee training programs: how to build a training system that delivers results

Employee training means any structured action an organization puts in place — internal or external — to develop transferable skills in the people who belong to it. It is an organizational function, not an event.

ISTAT data show that in 2020, 68.9% of Italian companies with at least 10 employees provided vocational training, a figure in line with the EU average of 67.4% [1][2]. The gap is not with Europe: it is internal, and it opens up below 50 employees (66.1% versus 95.5% of companies with more than 250) and above all in the share of employees actually involved in courses (30.4% versus 58.0%).

This pillar page draws the boundaries: what training is and what it is not, which skills are worth developing today, how to build a training plan that fits the size of your company, how to measure the return, and which common mistakes drain the investment.

Identify training needs before choosing the content

What evidence should a training plan start from: the market, department heads, or team members themselves? All three sources are partial. None of them, on its own, is enough to define a real need.

Many Italian companies choose the "what" (the course, the provider, the catalog) before the "why" (the missing skill, the operational problem). Eurostat data show that the main reason European companies give for not providing training is that "existing skills correspond to needs" (74.7% of cases) [2]: an implicit diagnosis, almost never verified. Training needs are identified by cross-referencing business objectives, skill gaps, and internal signals.

It helps to set a few working distinctions before going into detail. Training develops transferable skills that apply across different contexts (running a meeting, handling a conflict). Instruction (task training) is teaching a specific task tied to a tool, a procedure, or a machine (using new management software). Both are needed, but they answer different needs: instruction fixes a specific lack of competence, while training builds capabilities that remain even when the tool changes. Upskilling strengthens skills already present in the current role, raising their level; reskilling requalifies a person for a different role. Confusing these categories is the first design mistake: you end up asking a management training course to solve an operational instruction problem, or vice versa.

The first operational tool is a three-column skill gap analysis. For each role in the organization, you list the skills the role requires (column 1), the skills actually held by the team member in that role (column 2), and the gap between the two (column 3). The tool requires no dedicated software: a spreadsheet is enough. Its value lies not in the absolute precision of the assessment but in the overall visibility: which gaps are concentrated in a few people (individual risk), which are spread out (organizational risk), and which are negligible (not worth a training investment).

The second tool is cross-referencing three sources of evidence. Business objectives show where the company intends to go over the next 12-24 months and which skills become critical to that trajectory. Department heads (or, in very small companies, direct managers) read the gap between expected and observed performance and identify operational gaps. Team members themselves, in structured conversations or light questionnaires, express perceptions of their own limits that often anticipate signals not yet visible to others. None of the three sources is enough on its own: business objectives can be disconnected from operational reality, department heads can confuse skill problems with motivation problems, and team members can overestimate or underestimate their own abilities. Cross-referencing the three sources reduces the systematic errors of each.

The third tool is indirect internal signals. Recurring errors, systematic delays at certain steps, difficulty introducing a new tool, customer complaints about a specific area: these are indicators of missing skills that show up operationally before anyone names them. Most training needs in smaller companies emerge through crises (visible errors, lost customers, operational bottlenecks) rather than strategic anticipation: reducing this asymmetry is one of the goals of a good skill gap analysis.

Distinguish technical, soft, and digital skills in the plan

How much do soft skills really weigh compared with technical skills in determining a team member's performance? ISTAT data on the skill shortages reported by Italian companies put the two families almost on a par: technical and operational skills at 32.0%, the ability to work in a team at 31.2%, and problem-solving aptitude at 29.8% [1].

The skills companies invest in can be grouped into three families: sector-specific technical skills, soft skills (communication, problem solving, time management), and basic and advanced digital skills. The OECD Skills Outlook 2023 report insists precisely on the need to support the acquisition of "a broad range of skills, at different levels of proficiency" to cope with the ongoing transitions [3]. The proportion among the three changes with role, sector, and company size. Anyone who confuses the three families builds unbalanced training plans: too "soft" if technical skills are neglected, too "tool-driven" if soft skills are overlooked.

Sector-specific technical skills are those that enable a team member to perform their role in their market: for an accounting clerk, knowledge of tax rules and accounting principles; for a production technician, knowledge of machines and processes; for a salesperson, knowledge of the product and the target market. They are the most visible skills and the easiest to detect in job interviews. They develop mainly through direct on-the-job experience and specific technical training (sector courses, professional certifications, regulatory updates).

Soft skills are those that make the application of technical skills effective in real contexts: internal and external communication, problem solving, time management, collaboration, conflict management, leadership. ISTAT data document that soft skills are among the most lacking according to companies themselves: the ability to contribute to teamwork (31.2%) and problem-solving aptitude (29.8%) come right after technical and operational skills, ahead of managerial skills (23.3%) [1]. They are the most neglected skills in companies' training plans: often considered "innate" or "a matter of character," they can in fact be trained with structured interventions.

Digital skills split into two levels. Basic digital skills (using productivity tools, searching for information, communicating in digital environments) are now necessary for most roles, including non-technical ones. Advanced digital skills (data analysis, automation, cybersecurity, managing specific platforms) are concentrated in specialized roles. The distance between the basic and advanced levels is wide: confusing them leads to overestimating staff skills or to expecting basic courses to produce advanced results.

scroll the table →

The proportion among the three families in the training plan varies significantly by sector and size. In a traditional manufacturing company, technical skills tend to weigh more; in a professional services firm, soft skills are central; in a company with a strong digital component, the balance tilts toward the advanced level. The working principle is simple: balanced training produces more value than training concentrated on a single family, even when resources are limited.

Build a training plan that fits the size of your company

Is an effective training plan more like a calendar or a step-by-step journey? Probably neither: it is closer to an annual budget with renegotiable priorities.

A training plan for a company with 8 team members does not look like one for a company with 80, and neither mirrors that of a self-employed professional who trains themselves and three external collaborators. There are, however, four common elements: measurable learning objectives, mapping of participants, choice of formats (classroom, on-the-job, e-learning, shadowing), and calendar and budget. ISTAT data show that in 2020 training activities other than courses — on-the-job training, shadowing, conferences, distance self-learning — involved half of Italian companies, up ten points from 2015 [1]: formal training is not the only channel, and often not even the main one.

The first block is defining measurable learning objectives. A learning objective is not "do training on X." It is: "at the end of the program, participants will be able to do Y, assessable through Z." The difference sounds rhetorical but is substantial: without assessment criteria defined before the program, at the end you will not know whether it worked. Three to five measurable objectives per year are enough for most companies; beyond that, the plan scatters and no objective gets adequate attention.

The second block is mapping participants. For each learning objective, you define who will take part and how. Mapping is not an organizational detail: it is the most important strategic decision in the plan. Training new hires without involving established roles produces an asymmetric organization; training only established roles without thinking about new entrants produces stagnation; training only operational functions without involving people in positions of responsibility produces transfer gaps (classroom learning does not make it into processes). Balanced mapping spreads the investment across the whole structure.

The third block is choosing the formats. Four modes coexist: classroom (in person or virtual, with an instructor), on-the-job (learning during real work, with guidance), e-learning (self-learning on structured content), and shadowing (mentoring, coaching, peer learning). Each format suits specific types of skills: the classroom works well for conceptual knowledge, on-the-job for applied skills, e-learning for basic training and regulatory updates, shadowing for soft and leadership skills. An effective plan combines the four formats according to the nature of the content.

The fourth block is the calendar and budget. A realistic plan accounts for real time constraints (training hours are taken away from other activities) and budget constraints (including resources from joint training funds, which require advance planning). A useful practice: set training priorities quarterly, accepting that the plan can be renegotiated in light of unexpected events. A rigid plan breaks at the first operational surprise; a flexible plan keeps its direction while adjusting the timing.

For a company with 8 team members, the plan can be brief: one page with three annual learning objectives, an aggregate budget, a rough split among the four formats, and quarterly checkpoints. For a company with 80 team members, the plan needs more structure: objectives differentiated by function, a detailed budget by cost center, a more structured calendar, and roles dedicated to managing the plan. For the self-employed professional who trains themselves and their external collaborators, the plan applies at the individual scale: three personal development objectives per year, a proportionate budget, and a format mix made up mainly of formal training and learning on the job.

Read next: an operational model for building a training plan tailored to your company.

For a broader view of the management mechanisms that accompany the training plan, a complementary read is available in Business management.

Choose between internal, external, and on-the-job training

Should you rely on an authoritative external instructor or on a colleague who is less polished but closer to the real context? Skill transfer can be faster with the colleague, while legitimizing the change can be more solid with the outsider.

Three formats coexist: internal training (an expert colleague transfers skills to others), external training (an instructor or third-party provider), and on-the-job training (guided learning during real work). ISTAT 2020 data indicate that classroom instruction remains the most widespread mode (59.5% of Italian companies), but activities other than courses have reached 52.9%, with a weight that varies by sector and size [1]. The choice is not only economic: it depends on the type of skill, the speed of transfer, and the availability of internal experts.

Internal training has three main advantages. The first is relevance: whoever trains knows the company's real context, its customers, its processes, and adapts the content to the specific situation. The second is consolidating know-how: when an expert colleague transfers skills, the organization protects its knowledge from dependence on a single person. The third is low cost, at least on the surface. The main limitations are just as real: the colleague who trains is not necessarily a good trainer (knowing how to do something and knowing how to teach it are different skills), preparation time is taken away from other activities, and perceived authority may be lower than that of an external instructor, especially for soft skills.

External training has three specific advantages. The first is authority: an external instructor with documented experience legitimizes the message in a way a colleague could not. The second is the outside perspective: it brings examples and practices from other contexts into the organization, broadening its repertoire. The third is professional training expertise: an experienced instructor knows how to design a learning path, manage classroom dynamics, and calibrate content to participants' level. The limitations are cost (significantly higher than internal training), the generic nature of the content (rarely perfectly calibrated to the specific context), and the risk of reduced applicability if there is no bridge to daily operational reality.

On-the-job training is the most underrated mode in structured plans. It consists of learning during real work, guided by a more experienced colleague or a direct manager. The advantages are maximum relevance (you learn exactly what you need, on real materials, in real contexts) and the integration of learning and application. The limitations are dependence on the quality of the guide (a good performer is not automatically a good transmitter of skills) and the difficulty of protecting learning time from daily operational pressure.

scroll the table →

The 2x2 decision matrix — criticality of the skill (high/low) by availability of an internal expert (high/low) — provides practical guidance. When criticality is high and an internal expert is available, internal training is the most efficient choice. When criticality is high and no internal expert is available, external training is justified. When criticality is low, e-learning or on-the-job training are adequate, regardless of internal availability.

The practice documented by Italian companies suggests that combinations of the three formats are more effective than single formats in isolation: external training followed by internal shadowing transfers skills better than stand-alone external training, because the bridge to operational reality is covered.

Plan upskilling and reskilling across the employee life cycle

In a company with 10-15 employees, does it make sense to talk about reskilling, or is it a category that only concerns large companies? The most successful reskilling cases often involve small companies, where role rotation is structurally more frequent.

The OECD Skills Outlook explicitly links the green and digital transitions to the need for learning opportunities that respond to emerging labor market needs [3]. For smaller companies, this means planning periodic upskilling for team members who stay in their role and reskilling paths for those moving to different jobs, often as a result of automation or market changes. The employee life cycle — from onboarding to professional maturity — defines the key moments when training investment pays off most.

The employee life cycle in a smaller company unfolds in four phases, each with specific training intervention points.

The entry phase (first 6-12 months) requires a concentrated investment in instruction and basic training: knowledge of tools, understanding of processes, familiarity with customers, integration with the team. It is the phase with the highest training intensity per hour worked: the ratio of learning hours to production hours is high, but it is a necessary investment for the team member to become fully productive. For practical detail on onboarding in the first months, a complementary read is available in Employee onboarding: how to structure the integration of new people.

The consolidation phase (6 months-3 years) shifts the focus from instruction to upskilling. The team member knows the role but can strengthen their skills to handle more complex situations, take on more autonomy, and access a wider scope. This is the phase in which a structured individual development plan — even a single page — produces visible results: two or three growth objectives, six-monthly checkpoints, varied formats (external training on a specific topic, shadowing a senior colleague, taking part in a cross-functional project).

The maturity phase (3-7 years) is the most neglected moment in companies' training plans. The team member is autonomous, covers the role effectively, and tends to be considered "done." Yet this is the phase in which the risk of skill obsolescence is highest: what was learned in the early years is no longer enough, but the pressure to learn seems lower. A targeted investment in strategic upskilling (advanced digital skills, the ability to handle broader responsibilities, the ability to mentor younger colleagues) protects against gradual aging. Many voluntary departures in Italian companies come precisely from team members in the maturity phase, who look elsewhere for the development opportunities not offered internally.

The transition phase (variable) is the moment for reskilling: when the team member's role has to evolve significantly, either because of organizational changes (new technologies, process changes, reorganizations) or because of the team member's personal choices (moving to a different function, stepping up to a new role). Reskilling requires a larger investment than upskilling: it means building new skills, not strengthening existing ones. The most successful reskilling experiences combine structured formal training, intensive shadowing in the first months of the new role, and protected time for learning.

scroll the table →

Planning across the life cycle requires moving from a "training for whoever needs it now" logic to a "training along the team member's path" logic. Companies that adopt this logic reduce the risk of concentrating investment only on newcomers and forgetting the established team members who remain the backbone of the organization.

Read next: how to design upskilling paths for team members already in their roles.

Measure the return on training beyond classroom satisfaction

How long is it reasonable to wait before assessing whether a training program has really worked? The time frames given in this section — 6-12 months for the "behavior" level, 12-24 months for the "results" level — are a working criterion set by the editorial team: they establish when it makes sense to measure, and they are not durations found by a study.

Bank of Italy analyses of human capital place staff training among the factors that contribute to labor productivity in Italian companies [4]. It is an association, not causal proof: causality would need to be verified case by case, and other variables (quality of management, market context, technology choices) contribute significantly. Measuring the return still requires going beyond "participant happiness" and adopting a structured approach — the four-level model, on which there is extensive academic literature [5] — consistent with the learning nature of the investment.

The first level is reaction: participants' satisfaction at the end of the training program. It is measured with short questionnaires at the end of the course. It is the most superficial level, but it remains useful as a signal of organizational quality: a program on which participants consistently report negative reactions is less likely to produce learning. The limitation of this level is well documented: satisfaction does not reliably correlate with actual learning. A charismatic instructor can produce high satisfaction without transferring applicable skills.

The second level is learning: what participants have actually learned. It is measured with knowledge tests, practical cases, and simulations at the end of the program. It requires defining, before the program, which specific learning outcomes you want to produce. Measuring learning is more reliable than measuring reaction, but it remains limited: what people can do in the classroom is not automatically what they will do on the job.

The third level is behavior: what participants actually apply on the job in the months after training. It is measured through direct observation by the manager, structured self-assessment, and feedback from colleagues and customers. It requires a time frame of 6-12 months to be reliable. It is more informative than the previous levels: it measures the real return on training, because only changed behaviors produce operational results.

The fourth level is results: the impact of training on the company's operational results (fewer errors, better quality, shorter lead times, higher customer satisfaction, higher productivity). It requires a time frame of 12-24 months and the ability to isolate the effect of training from other variables that affect the same indicators. It is the most relevant level for the investment decision, but also the hardest to measure rigorously.

Consider the hypothetical case of a company that invested in training on running effective meetings. The reaction level is measured at the end of the course (satisfaction questionnaire). The learning level is measured with a practical test at the end of the program (e.g., designing a meeting agenda and simulating how to run it, assessed against written criteria). The behavior level is measured at 6 months: do the company's meetings have structured agendas on average? Has the average duration changed? Has participants' perception of effectiveness increased? The results level is measured at 12-18 months: has the organization's decision-making efficiency improved? Has the time spent in meetings decreased for the same output? Have side effects emerged (on team members, customers, processes)?

A realistic practice for smaller companies is to measure the first two levels (reaction and learning) systematically, and levels three and four on a selection of the most important training programs, where the investment justifies the measurement effort. Measuring every program at all four levels is not organizationally sustainable in companies with fewer than 100 employees.

The methodological caveat remains central: even the most rigorous measurements at levels three and four show correlations, not direct causation. Training is one variable among many that influence behaviors and results. The practical guidance offered here should be read as orientation, not as unambiguous proof of impact.

Involve middle managers in transferring skills

How much of the responsibility for a failed training program lies with the instructor, and how much with the manager who should support its application? The available data suggest that responsibility for transfer lies much more with the context of application than with the classroom.

Training that stays in the classroom and does not make it into processes produces no change. The meta-analysis by Blume and colleagues of 89 studies places support from the direct supervisor — department head, functional manager — among the contextual factors associated with the transfer of learned skills, with a correlation of .31 versus .14 for peer support, and with the caution the authors themselves express about the small number of samples available [7]. If the direct supervisor does not support application, training remains an isolated event. This holds both in small companies where the "middle manager" is the business owner, and in more structured organizations.

The middle manager's role in transfer plays out at three moments: before, during, and after training.

Before training, the middle manager has three responsibilities. The first is making clear to the team member why they will take part in the program: which skill is being developed, how it connects to the role, what is expected on their return. The second is protecting training time: training interrupted by calls, emergencies, and requests from the manager tells the team member that training ranks below ordinary work. The third is preparing the context: lightening workloads during training weeks, letting the team know the person will be less available, and avoiding a pile-up of deadlines that will create pressure on their return.

During training, the middle manager has a smaller but not negligible responsibility: showing interest, asking how it is going, and avoiding interruptions except for real emergencies. The behavioral signal (this training matters to us) often carries more weight than the content of the training itself.

After training, the middle manager has the most important responsibilities. The first is dedicating structured time to the return: a 30-45 minute conversation in which the team member describes what they learned, what they want to apply, and what obstacles they foresee. The second is creating opportunities for application: assigning the team member concrete tasks in which they can put into practice what they learned, even within a controlled scope. The third is giving specific feedback on changed behaviors: recognizing when the team member applies the new skills and pointing out when they fall back into old habits. The fourth is assessing over time: 3-6 months after training, a structured conversation about actual changes and necessary adjustments consolidates learning.

A short checklist for middle managers includes five essential actions: communicate the purpose of the training to the team member before it starts; protect training time from interruptions; hold a return conversation within the first week; create at least one structured opportunity for application within a month; check how behaviors have evolved at 3-6 months.

Read next: how to develop the skills of middle managers who need to support training transfer.

When the middle manager is the business owner (typical in companies with fewer than 15 employees), the five actions on the checklist do not lose weight: if anything, they matter even more, because the business owner's signal is the organization's signal. Neglecting transfer in these situations is the fastest way to drain your training investment. For more on the middle manager's role in smaller companies, a complementary read on writing the job description is available in Job description.

Integrate structured training and informal learning on the job

If informal learning accounts for the dominant share, does it still make sense to invest in structured training? Yes, because without a structured intervention informal learning tends to reproduce existing habits rather than change them.

The review published in the International Journal of Training and Development indicates that a large majority of adult learning in companies happens informally: by observing, making mistakes, asking a colleague [5]. The figure is an indicative estimate, not a general law, and it varies significantly by sector and role. Structured training becomes more effective when it is designed to trigger and accelerate that informal learning rather than replace it. The companies that get the best results do not necessarily have more classroom hours: they often have an "infrastructure" of shadowing, mentoring, and feedback that the formal plan feeds.

The 70-20-10 model — 70% learning from direct experience, 20% from interactions with colleagues and mentors, 10% from formal training — is a useful summary for rebalancing attention, even though it remains an indicative estimate with limited empirical grounding. The model's value lies in correcting the widespread tendency to concentrate investment and attention only on the formal 10% (courses and budget) while neglecting the 70% and the 20%.

The principles of effective adult learning, applied here to the workplace, are covered in depth in Effective adult learning.

The 70% (direct experience) requires active design, not just waiting for people to "learn by doing." Learning from experience is greatest when assigned tasks become progressively more complex (relative to the team member's current level), when there are protected spaces for mistakes (reversible, contained), and when the team member has time to reflect on what went well and what did not. Without these conditions, experience reproduces existing habits — and skills stay flat over time. A useful practice is the structured debrief after significant events: 30 minutes of explicit conversation about what happened, what was learned, and what will be done differently next time.

The 20% (interactions with colleagues and mentors) needs structure so it does not stay random. Three mechanisms work in smaller companies: structured peer learning, in which team members with different skills exchange knowledge in specific areas (one hour a week is already significant); mentoring, in which a senior team member supports the development of a less experienced one on a specific skill for a defined period (3-6 months); and external professional communities, in which the company's team members exchange ideas with peers from other organizations (trade associations, industry events, professional networks). These practices cost little and have a disproportionate impact.

The 10% (formal training) is no less important for being the quantitative minority. It is the component that provides frameworks, theoretical models, a shared vocabulary, and exposure to experts — elements that informal learning can hardly produce. Formal training is effective when it is designed to trigger the two previous levels: a leadership course not followed by concrete opportunities for application and peer learning among participants remains an isolated event. The structured training plan that gets the best results combines formal training (10%) with targeted interventions on the 70% and the 20%, building a complete infrastructure.

The working principle is not to replace informal learning with structured training, but to amplify it. Formal training that works is training that activates informal learning: it raises questions people keep carrying with them, provides frameworks that apply in concrete situations, and creates moments of exchange that continue in the following months. For more on the link between training and organizational change, a complementary read is available in Organizational change management.

Common mistakes that drain the investment in training

Of the six mistakes, which one costs a company with 10-50 employees the most over the medium term? Almost always the fifth: not involving the middle manager. Without them, learned skills do not become practice.

The INAPP report on continuing training in Italy identifies as a fundamental weakness "the limited spread of a culture oriented toward the continuous development of skills, especially in small enterprises, where the resources and structures dedicated to training are insufficient" [6]. Six mistakes recur more often than others: choosing courses before analyzing needs, training only new hires while forgetting established roles, separating technical training from management training, measuring only satisfaction, leaving the middle manager out of the process, and treating training as an event rather than a continuous function. Recognizing them avoids spending budget without moving real skills.

The first mistake is choosing courses before analyzing needs. You start from the provider's catalog or an event brochure and pick what "might be useful." The frequent result is investing in partially relevant topics while leaving the real gaps uncovered. The alternative practice: complete the skill gap analysis before looking at the training on offer. Needs drive the choice, not the other way around.

The second mistake is training only new hires while forgetting established roles. It is a mistake of asymmetry: resources are concentrated on onboarding, because the need is visible and immediate, and the ongoing development of senior team members is neglected. The medium-term result is the gradual obsolescence of skills in the organization's core roles, and the loss of the most capable team members, who look elsewhere for the development opportunities they did not find internally. The alternative practice: in the training plan, spread the investment evenly across the phases of the employee life cycle.

The third mistake is separating technical training from management training. Training that develops only technical skills without addressing soft skills produces specialists who struggle to coordinate; training that develops only management skills with no connection to technical content produces managers disconnected from operational reality. The alternative practice: design programs that integrate technical and soft content, even if different instructors handle the components, but linked within the same plan.

The fourth mistake is measuring only satisfaction. The "reaction" level is the easiest to capture but the least informative. When the only metric used is participant satisfaction, the organization does not know whether training actually produced learning, behavior change, or operational results. The alternative practice: add at least the "learning" level (checking the skills acquired at the end of the course) for all programs, and the "behavior" and "results" levels for the most important ones.

The fifth mistake is leaving the middle manager out of the process. It is the mistake with the highest hidden cost: even well-designed training dissolves if the participant's direct supervisor does not support its application. The early warning sign is training that "doesn't show" in the participant's behavior 2-3 months after their return. The alternative practice: involve the middle manager in all three phases (before, during, after), using the short five-action checklist described in the dedicated section.

The sixth mistake is treating training as an event rather than a continuous function. A training program is launched in response to a specific problem (a visible error, a lost customer, an operational difficulty), and then things go back to normal. The result is swinging between periods of intense training and periods of no training at all, with no continuous development trajectory. The alternative practice: build the training plan into the company's ordinary routines (annual planning, six-monthly reviews, individual development conversations), protecting budget and time as you would for any other organizational function.

The six mistakes are not independent: they reinforce one another. A skipped needs analysis produces poorly chosen courses, delivered to poorly selected participants, separated by type, assessed only on satisfaction, with no involvement of the middle manager, in an isolated-event logic. Tackling a single mistake without addressing the others produces limited results.

Limitations and conditions of applicability

The guidance in this guide was written for Italian companies ranging from the self-employed professional with one to three team members to the company with 100 employees. Larger companies operate in contexts that require structured HR functions and more elaborate training management mechanisms, which this pillar page does not cover.

The data cited come from Italian and European sources [1][2][3][4][6]. Transferring them to non-European contexts requires caution: training practices are strongly shaped by the national regulatory system (joint training funds, collective agreements, public funding for continuing training) and by organizational culture, and models that work in Italy may need significant adaptation elsewhere.

The documented associations between training investment and performance indicators (productivity, quality, retention) [4] do not imply direct causation. A company's performance depends on many variables — sector, quality of management, market context, product quality — of which training is only one. The practical guidance offered here should be read as orientation, not as a quantitative guarantee.

The 70-20-10 model and the estimate of the share of informal learning in companies [5] are orders of magnitude, not general laws. The actual proportions vary by sector, by role, and by phase of the employee life cycle. The model's value lies in rebalancing attention, not in the numerical precision of the proportions.

The four-level model for evaluating the return on training requires methodological skills in data collection and interpretation that not every company has in-house. A simplified version, which systematically measures the first two levels and selects the most important programs for levels three and four, is the realistic practice for companies with fewer than 50 employees.

Finally, working on training without acting on the other dimensions of the organization (people management, leadership, processes) produces limited results. Training is a cross-cutting function: it is reinforced by consistent interventions in hiring, performance review, development, and coordination, and weakened when these areas pull in different directions.

FAQ

How much should a company invest in training, as a percentage of labor costs? European statistics document significant variability across sectors and countries: in 2020 the share of companies providing training ranged from 63.5% of small enterprises to 92.8% of large ones on the EU average [2]. In the same CVTS survey, the cost of courses as a share of labor costs in Italy rose from 1.3% in 2015 to 1.7% in 2020, above the EU-27 average of 1.5% [2]. More than the percentage, what matters is the internal consistency of the plan: it is better to invest a smaller share in a structured way than to scatter a larger share across interventions unconnected to real needs.

Can effective training be done in a company of 5 people? Yes, with two main adaptations: the plan should be sized at the individual scale (personal development objectives for each team member rather than group programs), and formats should lean toward modes with a low organizational cost (e-learning for standard content, shadowing for applied skills, targeted external training for significant leaps). The principle that training is a function, not an event, applies at any size.

How do you fund training with limited resources? Three levers coexist. Joint training funds (managed by employer and employee representatives, where your country has them) cover a significant part of the cost of formal training; they require advance planning and membership in the fund. Public calls for funding finance specific programs, often tied to priority topics (digitalization, sustainability, safety). Internal training and structured informal learning have low direct costs, even though they require time from competent internal people.

Is it worth training staff given the risk that they will leave the company? The dilemma is the one raised in the opening question. Bank of Italy analyses of human capital [4] place training among the factors that contribute to company productivity: the cost of not training (skill obsolescence, reduced ability to adapt, lower quality of work) is generally higher than the cost of the risk of losing trained team members. Documented practice also suggests that companies that invest in training tend to retain their team members better, because professional development is one of the most important drivers of satisfaction.

How long should you wait before assessing the effectiveness of a training program? The "reaction" level is measured immediately; the "learning" level at the end of the course or in the following weeks; the "behavior" level needs 6-12 months to be reliable; the "results" level needs 12-24 months and the ability to isolate the effects of training from other variables. A reliable assessment of the overall return therefore requires a time frame of at least 12 months.

How do you integrate mandatory training with strategic training in the same plan?

Mandatory training — workplace safety, sector regulatory updates — already takes up part of the annual calendar, whether or not there is a structured plan.

The INAPP report on continuing training [6] notes that, in Italy, joint training funds reach about 20% of workers and 12% of member companies each year: at that pace, a worker gets a funded training opportunity every five years and a company every eight, with even less favorable proportions for micro-enterprises.

The most efficient way to integrate the two components is not to treat them as separate chapters, but to use the mandatory training calendar as the base onto which strategic priorities are grafted.

If a regulatory update already requires a day of classroom training for a group of team members, the same session can host an additional module on a priority skill, at a modest incremental cost compared with organizing two separate events.

This integration requires knowing the calendar of regulatory obligations in advance, information that in most companies is already available from whoever manages safety or sector compliance.

The training plan that fits the size of your company is the natural place to plan this integration.

Can peer learning replace formal courses in a smaller company?

Peer learning does not replace formal courses: it complements them, and does so effectively only under certain conditions.

The review on formal and informal learning in the workplace [5] shows that soft skills develop largely through practice and exposure to real contexts, not only through structured programs — evidence that supports the value of peer learning, without making it equivalent to a structured program.

Peer learning works when it has a minimum structure: a defined topic, protected time (even just one hour a week), and an explicit mode of exchange, not informal conversations left to chance.

Without these elements, peer learning tends to reproduce the knowledge already present in the group, without introducing anything new.

Formal courses remain necessary for content that requires an external framework, a shared vocabulary, or exposure to practices not yet present in the company: no colleague, however expert, can transfer a skill the group does not already have.

The most effective combination has formal training provide the conceptual frame and peer learning support its application in the following months.

Operational summary

Employee training is a continuous organizational function, not an event. It is organized into six connected areas: identifying needs, selecting priority skills, building the plan, choosing formats, integrating training across the employee life cycle, and measuring the return. The working principle running through all six areas is simple: every training intervention must be linked to a measurable objective, an identified participant, and a mechanism for transfer into daily practice.

The starting point is diagnostic: cross-reference business objectives, skill gaps identified by department heads, and internal signals from team members, distinguishing the three families of skills (technical, soft, digital). The second step is design: build a four-block plan (measurable objectives, participant mapping, choice of formats, calendar and budget) that fits the size of the company. The third step is execution: combine internal, external, and on-the-job training according to the type of skill, integrating upskilling and reskilling across the employee life cycle.

Measuring the return requires going beyond the "reaction" level and adopting a structured four-level approach (reaction, learning, behavior, results), with the awareness that you are measuring correlations, not direct causation. Middle manager support is among the contextual factors that the meta-analysis on training transfer most clearly associates with applying what has been learned [7]: without their involvement before, during, and after, even well-designed programs risk remaining isolated events. Integrating structured training and informal on-the-job learning (the 70-20-10 model) is the most effective practice in smaller companies: formal training that works is training that amplifies informal learning, not training that claims to replace it. The six recurring mistakes — choosing courses before needs, training only new hires, separating technical and soft skills, measuring only satisfaction, excluding the middle manager, treating training as an event — are structural patterns that can be corrected with targeted interventions.

Conclusion

Building a training system in a company means turning an episodic activity — the course, the provider, the catalog — into a continuous organizational function: identifying needs, choosing among formats, integrating with informal learning, involving middle managers, and measuring the return beyond classroom satisfaction. Companies that apply this approach do not necessarily have more training hours: they have better-directed ones.

Training is one of the three pillars of people management in a company, together with onboarding and role definition. To explore related topics, also read Employee onboarding: how to structure the integration of new people and Business management: the operational pillars.

If Italian companies with fewer than 50 employees closed the internal gap — almost thirty percentage points behind larger companies in the rate of training provision, and almost as many in the share of employees involved [1] — the human capital of Italy's productive system would come out stronger [4]. It is not a promise; it is a direction of work.

Sources and references

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

[2] Eurostat, "Statistics on continuing vocational training in enterprises", Statistics Explained, 2022. Available at: https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Statistics_on_continuing_vocational_training_in_enterprises

[3] OECD, "OECD Skills Outlook 2023: Skills for a Resilient Green and Digital Transition", OECD Publishing, Paris, November 6, 2023. Available at: https://www.oecd.org/en/publications/oecd-skills-outlook-2023_27452f29-en.html

[4] Banca d'Italia, "Relazione annuale", Banca d'Italia, Rome. Available at: https://www.bancaditalia.it/pubblicazioni/relazione-annuale/

[5] Manuti A., Pastore S., Scardigno A. F., Giancaspro M. L., Morciano D., "Formal and informal learning in the workplace: a research review", International Journal of Training and Development, 19(1), 2015. Available at: https://doi.org/10.1111/ijtd.12044

[6] INAPP, "XXIV Rapporto sulla Formazione Continua in Italia — Annualità 2023-2024", Istituto Nazionale per l'Analisi delle Politiche Pubbliche for the Italian Ministry of Labor and Social Policies, 2025. Available at: https://oa.inapp.gov.it/server/api/core/bitstreams/75279f91-963e-4b73-8bd1-52bc00177be7/content

[7] Blume B. D., Ford J. K., Baldwin T. T., Huang J. L., "Transfer of Training: A Meta-Analytic Review", Journal of Management, 36(4), 2010, pp. 1065-1105. Meta-analysis of 89 empirical studies: transfer climate shows the strongest relationship with transfer (.27), followed by support (.21); in the subsequent analysis, supervisor support (.31) is more closely related to transfer than peer support (.14), on samples the authors themselves describe as small. Available at: https://doi.org/10.1177/0149206309352880