On-the-job training form: a template to track skills and decide on the next steps

A useful on-the-job training form answers three questions: which skill is targeted, what was observed in the actual work situation, and what decision follows. When the goal is to track skills development, the other information (dates, length of sessions, topics covered) should help answer these three questions.
Some mentoring forms are built mainly around meetings, activities completed and topics covered. This information is useful for following how the mentoring unfolds. However, when a company wants to know whether an employee has mastered the skills their job requires, it needs to be complemented by observations of the work actually performed.
The form then serves two people. The manager finds what they need to decide: assign a task, plan additional training, request a validation. The employee sees their progress, what they still need to develop, and skills acquired on the job that would often remain informal without such a record.
This article first places the form within the process it belongs to. It then proposes seven sections, a scale that distinguishes the support provided from the conformity of the work, a template, a worked example, guidelines for setting the frequency of check-ins and a method for aligning assessments across mentors. The form remains an individual tool: managing the whole programme (hours, indicators, multiple sites) falls under tracking workplace mentoring.
Key takeaways
An on-the-job training form records what an employee does in a real work situation, how they do it, and the decision that follows.
It requires preparation: expected skills, target level, criteria and work situations.
Performing a task without help does not mean performing it correctly. The rating scale must therefore account for both the support needed and conformity with the criteria.
How often check-ins take place depends on the job, the risks involved and the opportunities to practise.
The form records, it does not validate on its own: validation requires relevant, sufficient evidence and an authorised assessor.
The form is not where the process starts
In a workplace mentoring programme, the training form assumes the company already knows what the employee needs to develop. Yet this is often the work that remains to be done. Without it, the form gets filled in, but the mentor does not know what to relate their observations to.
Before building the form, you therefore need a competency framework that sets out the skills expected in the job, the level of mastery expected for each one and the criteria used to assess that mastery. You also need to have identified the work situations in which these skills can be developed and assessed. The form comes in afterwards, at a specific point in the process:
Identify the required skills: what does the employee need to be able to do?
Define the criteria and expected levels: how do you recognise that a skill has been mastered?
Assess the employee's starting point: which skills do they already have?
Organise learning situations: what do they need to practise, and with what support?
Track observations and progress: what change can be seen? This is the main scope of the form.
Assess and decide: what can be validated, and what remains to be developed? The form contributes to this step without replacing it.
Each step relies on a different tool. Confusing them leads to asking the training form to do what it cannot do on its own.
Tool | Main purpose |
Competency framework | Define the expected skills |
Initial assessment | Identify skills already mastered and gaps |
Development plan | Organise learning |
On-the-job training form | Record observations, progress and decisions |
Assessment grid | Assess mastery against criteria |
Skills portfolio | Capture the skills developed and assessed |
A form built around meetings shows how mentoring unfolds, not what has been mastered
A training form can record four different things: what was planned, what was done, what was assessed, and how the result is formalised. A form organised around dates, topics covered and objectives for the next session covers the first two well. It says much less about the third, which is precisely what a manager needs before assigning someone to a job.
What a meeting-based form can show | What it cannot show on its own |
The employee was mentored four times in three weeks. | They have mastered the tasks covered. |
They were shown and taught how to lock out a machine. | They can carry out a lockout without missing a step. |
The mentor feels "it's going well". | Another mentor would reach the same judgement. |
The form is signed by both parties. | The skill is validated. |
Such a form is a self-reported record of how the mentoring unfolded, and its value depends on the information and evidence available. To decide on autonomy, an assignment or additional training, it must be complemented by what was observed, on which task, under which conditions and with what result. This is what turns a meeting log into a skills record.
The seven sections we propose
A useful form must remain simple enough for mentors to fill in regularly. We propose seven sections, each with a role in the final decision. This is a methodological proposal: each company can adapt it to its jobs and constraints.
1. Identification: who, in which job, over what period
Employee's name, mentor's name, job, start date in the job, date of the check-in. One precaution: record the job actually held rather than the contract title. A "production operator" can hold very different jobs depending on the workshop.
2. The target skill, described by observable criteria
This is the section that makes the difference. "Knows the lockout procedure" cannot be observed. "Carries out the lockout in the right order, fits their padlock and checks for zero energy before starting work" can.
If two mentors watching the same scene do not reach the same conclusion, the criterion may be poorly worded. The disagreement may also come from a different understanding of the criterion, a lack of mentor preparation, or what each of them actually saw. It is worth finding the cause before rewriting the criterion (see the calibration method below).
The number of skills covered at each check-in depends on the job, the length of the mentoring period and the opportunities to practise. It is generally better to focus on the skills that determine readiness for the job than to try to cover the whole framework every time.
3. The work situation actually observed
Record the task, the conditions (alone or in a team, normal or reduced pace, routine or incident) and what the mentor saw. A skill demonstrated on a stopped machine, at the start of a quiet shift, says little about what will happen at the end of a night shift with a breakdown. The conditions of the observation limit what can be concluded from it.
4. The mastery level: distinguishing support from conformity
This is the most delicate part of the form. A scale based solely on the mentor's involvement (done with them, under observation, alone) describes how much support was given. It does not say whether the task was done correctly. And a person can work alone and produce a result that does not conform.
Observed situation | Autonomy | Demonstrated mastery |
The employee performs the task correctly with instructions at every step. | Low | Partial |
The employee performs the task alone, but the result does not conform. | Apparently high | Insufficient |
The employee performs the task correctly under observation, without any help. | Potentially high | Potentially high |
The employee performs a conforming task alone, in various representative conditions. | High | Supported by observations |
We therefore propose a scale that combines both dimensions. It is a methodological proposal, not an official framework.
Level | Description |
Not assessed | There is not enough evidence to reach a judgement. |
1. Learning | The person does not yet perform the task in line with the criteria, despite the support provided. |
2. Mastery with support | The person performs the task in line with the criteria, with the mentor's help. |
3. Mastery in an observed situation | The person performs the task in line with the criteria, without support, in a given situation. The mentor observes without intervening. |
4. Confirmed mastery | The person consistently performs the task in line with the criteria and without support, in the representative conditions required for the skill. |
This has two practical consequences. First, for some skills the final result does not show whether the critical steps were followed: lockout is a case in point. Level 3 then requires the mentor to have observed the work itself, not just checked the result. Second, "Not assessed" is not a failure: it means no situation has yet made it possible to observe the skill.
5. The mentor's feedback and the employee's view
Two separate areas. The mentor writes what they saw and what still needs work; the employee writes what they feel they have acquired and what still causes them difficulty. When the two readings differ, for example an employee who feels confident about a task the mentor considers fragile, the discussion can help identify a difficulty that neither would have put into words alone.
This is also the section that involves the employee in analysing their own progress. They do not simply receive an assessment: they see what they have already acquired and what they still need to develop.
6. The evidence supporting the assessment
An assessment whose supporting observations are not made explicit is hard to share and to verify. Depending on the trade, this evidence may be a dated observation, a part produced, a quality control record or a photo of an assembly. The aim is not to pile up documents, but to be able to answer a simple question later: what was this judgement based on?
7. The decision and the next step
Each form should end with a decision: continue in the same situation, move to a more demanding situation, request a validation, or flag a difficulty to the manager. A form that leads to no decision is likely to remain a simple report.
An important point: the mastery level alone is not enough to authorise someone to work on their own. Site safety rules, required training and authorisations, and lone working conditions are also part of the decision.
Three questions to ask before filing a form
Without knowing the context, can you tell which skill was being worked on?
Would another mentor, facing the same situation, have given the same level?
Does the decision follow from what was observed, or from a general impression?
When a skill is not mastered, the next step depends on the cause
"Continue in the same situation" is the right decision in many cases, but it does not say how. Two employees at the same level may need very different actions, because their difficulty does not have the same origin. The mentor therefore benefits from identifying the cause before choosing what comes next.
Observed difficulty | Possible learning action |
The employee does not understand the procedure. | Explanation, demonstration, then a check of understanding |
They know the procedure but cannot apply it. | Supported practice and regular feedback |
They handle the usual situation, but not unusual ones. | Varied practice situations, in a suitable setting |
They have not yet had the opportunity to practise. | Planning new work situations |
They keep making the same mistake. | Analysing the mistake with them and adapting the support |
This grid does not claim to be exhaustive. Demonstration, supported practice and varied practice situations correspond to different work-based learning methods, which do not address the same difficulties. The grid is mainly there to turn the form from a tool that records progress into a tool that organises what comes next. The "Next step" column of the template is designed for this.
The on-the-job training form template
The template covers the seven sections. It is filled in after the observation rather than during it: the mentor completes their part, the employee adds theirs, and the manager reads the decision. How long it takes depends on the number of skills covered; what matters is that it remains compatible with the mentor's workload.
Header
Section | To fill in |
Employee | Name, job actually held, start date in the job |
Mentor | Name, role |
Check-in | Date, check-in number |
Skills observed
Target skill | Observable criteria | Situation observed | Level (Not assessed, 1 to 4) | Supporting evidence | Mentor's comments |
Employee's view
What I feel I have acquired | What still causes me difficulty |
Decision
Skill | Decision | Cause identified (if not mastered) | Next step | Deadline |
Continue / more demanding situation / request validation / difficulty flagged to manager |
Example: a maintenance technician's form at the third check-in
Illustrative example. Lockout rules, required authorisations and lone working arrangements depend on each site's procedures.
A maintenance technician joined a production site two months ago. Their mentor, an experienced technician in the team, holds the third check-in on two skills required before any intervention without a colleague.
Target skill | Observable criteria | Situation observed | Level | Supporting evidence | Mentor's comments |
Lock out equipment before working on it | Follows the procedure in order, fits their padlock, checks for zero energy before starting work | Tool change on a press, normal pace, day shift. The mentor observes the work without intervening. | 3. Mastery in an observed situation | Direct observation on the day, completed lockout sheet | All steps followed in order, zero-energy check done without prompting. Observed on a single piece of equipment, during the day. |
Diagnose a conveyor breakdown | Starts from the symptoms, consults the documentation, checks simple causes before dismantling | Conveyor stoppage, sensor fault | 1. Learning | Observation on the day | Dismantled the sensor before checking its power supply, despite a hint from the mentor. A diagnostic approach is not yet in place. |
Employee's view: "Confident with lockout. With breakdowns, I don't know where to start when there are several possible causes."
Decision:
Lockout: mastery has been observed in a single situation. Before requesting a validation, further observations are planned on other equipment and in other conditions. Working alone will also depend on the required authorisations and the site's lone working rules.
Diagnosis: the cause identified is a lack of method rather than a lack of technical knowledge. The mentor plans to demonstrate their own diagnostic approach, followed by supported practice on two simple breakdowns before the next check-in.
The same form therefore leads to two different decisions and two tailored actions. A meeting-based form might have recorded "check-in done, good progress". It would not have shown that the technician is making progress on lockout, that they have not yet mastered diagnosis, or why.
How often to hold check-ins depends on the situation
Requiring a full form after every exchange risks overloading the mentor and discouraging its use. Formal check-ins can therefore be spaced out, provided they take place when a decision is being made. Between two check-ins, the mentor simply notes significant observations (a mistake, a success in a difficult situation), which will serve as supporting evidence.
No single frequency works everywhere. Rather than one fixed schedule, we propose criteria for building your own:
Situation | Suggested frequency |
New starter in the job | Frequent check-ins at the start |
Skill involving significant risk | Observations matched to the risks and validation requirements |
Skill used daily | Tracking based on real opportunities to practise |
Skill rarely used | Tracking triggered as soon as a suitable situation arises |
Employee experiencing difficulties | Temporarily more frequent check-ins |
Skill whose mastery is established | Less frequent tracking, or periodic checks as needed |
This table is a methodological proposal, not a standard. As an example, here is what it might look like for a job onboarding of around six months:
When | Who takes part | What is filled in | Expected decision |
Before the start date | Manager, mentor | Target skills and criteria, pre-filled from the initial assessment | Choose the first work situations |
End of the first week | Mentor, employee | First observations | Confirm or adjust the situations |
End of the first month | Mentor, employee, manager | Review of priority skills | Which tasks can be assigned, and with what support |
Third month | Mentor, employee | Levels reached on each priority skill | First validation requests |
End of the period | Mentor, manager, authorised assessor | Summary of all target skills | Mastery validated, or further development planned |
A shared scale is not enough: mentors also need calibrating
A shared scale is necessary. Without a common practice, however, different mentors may interpret it differently. The Australian Skills Quality Authority (ASQA) treats this as a principle in its own right, reliability: evidence is interpreted consistently, and the outcome does not depend on who carries out the assessment.
To get closer to this, an L&D manager can regularly run a simple calibration exercise:
Choose a few representative skills, preferably those that determine readiness for the job.
Ask several mentors to assess the same situation, live or from a video.
Compare their assessments.
Identify the reasons for any differences: an ambiguous criterion, a different reading of the scale, or evidence seen by one mentor and not the other.
Clarify the criteria, or add an example to them.
Repeat the exercise, especially when new mentors join the programme.
The exercise takes time. In return, it addresses a problem the scale alone does not solve, and it develops mentors in their role as assessors. It matters even more in multi-site skills assessment, where each site often ends up building its own benchmarks.
Five mistakes that make a form hard to use
A training form is only useful if mentors fill it in, and a mentor is first and foremost an operational employee with production targets to meet. Most difficulties stem from this.
Reusing a template designed for another purpose. A template built to follow how mentoring unfolds does not necessarily show whether a skill has been mastered on the job. Check that it answers the three starting questions.
Writing vague criteria. "Good command", "appropriate attitude" or "sufficient knowledge" leave each mentor free to interpret, and the forms become hard to compare.
Filling in the form in front of the employee, without discussion. The form may then feel like an inspection, and the employee may keep quiet about what they have not mastered, even though this is exactly the information the mentor needs.
Letting each mentor use their own scale. Assessments can no longer be compared from one mentor to another. A shared scale is the first condition; calibration is the second.
Filling in forms that are never used. If the forms never lead to a decision, mentors may fill them in as a formality, then stop filling them in altogether.
A training form is not enough to validate a skill
The form is a record. Validating a skill is a decision, made by a person authorised to do so, which requires a skills assessment based on relevant and sufficient evidence. A series of well-completed check-ins does not, on its own, amount to validation.
To judge the quality of the evidence an assessment relies on, you can draw on the four rules of evidence applied by ASQA (Practice Guide: Assessment, 2025). These rules belong to the Australian vocational training framework and are not a regulatory requirement in France or elsewhere.
They do, however, offer a useful lens for deciding whether the observations gathered really allow a conclusion :
Validity: the evidence makes it possible to assess the target skill, not a neighbouring task.
Sufficiency: the quality, quantity and relevance of the evidence support an informed decision. The number of observations matters, but is not enough.
Authenticity: the evidence reflects the work of the person being assessed, not that of their colleague.
Currency: the evidence shows skills mastered today. How long an observation remains representative depends on the skill and on how practices evolve.
These rules assume, once again, that mentors share the same grid and read it in the same way. More broadly, they inform tracking learning at work, where the question is whether on-the-job support has actually produced a skill.
This is the approach taken at Saur, the French water and sanitation group, for its sewer jetting truck drivers, who work on wastewater networks with precise technical procedures and strict safety constraints. External training courses provide the fundamentals of the trade.
They do not necessarily cover all the practices, procedures and work situations specific to the company. With Daylindo, the challenge was therefore to organise the transfer of know-how specific to Saur, described with its in-house experts, and then its assessment in real work situations before any independent intervention.
From paper to oversight: what changes as the number of employees grows
As the number of employees being mentored grows, the challenge is no longer just collecting observations, but consolidating and using them. Twenty mentors each supporting two people, with one check-in a month, produce for example nearly 500 forms a year.
Paper or spreadsheet tracking can work on a small scale. However, it becomes harder to keep an up-to-date skills mapping of what is actually mastered, to follow the gaps still to be addressed and the time needed to reach the expected level in each job. This last indicator, time to proficiency, can be read by job, by site or by mentor. It should be interpreted with care: a long delay may come from the employee, but also from a lack of formative work situations or a mentor with little availability.
This is what Daylindo, a frontline skills management platform, makes possible. Field trainers record their observations at the workstation, from a mobile device. Each validated skill is recorded with the evidence that supports it. Managers can see by site, team and employee who has mastered what, and employees keep a record of the skills they have developed. The platform does not replace the field trainer. It makes their tracking visible, traceable and demonstrable.
A skills record rather than a meeting log
An on-the-job training form is not only there to show that mentoring took place. It should show what the employee does, under which conditions and with what result, and then support a decision on what comes next. This requires groundwork on the expected skills, a scale that distinguishes support from conformity, mentors who read it in the same way, and a decision at every check-in. For the employee, it is also what makes skills acquired on the job visible.
FAQ
Who should fill in the on-the-job training form?
The mentor fills in the observation, level and decision sections. The employee completes their own area on what they feel they have or have not acquired. The manager reads the decision and steps in when a difficulty is flagged.
How often should a training form be completed?
There is no single frequency. It depends on the stage (check-ins are generally more frequent at the start), the level of risk of the skill, the opportunities to practise and any difficulties the employee encounters.
What is the difference between a training form, a training logbook and an assessment grid?
The training form records a check-in: what was observed and what was decided. The logbook brings together all the forms and documents over the whole mentoring period. The assessment grid is used to judge a level of mastery against defined criteria. It can be used for the initial assessment, during the mentoring period to track progress, or at the time of validation.
Should the employee sign the form?
It depends on the context. For internal mentoring, a signature mainly shows that the employee has read the form and been able to give their view. Within a regulated scheme (apprenticeship, funded training), refer to the documents required by the organisation concerned.
Can the same form be used for an apprentice and a new employee?
The structure can be the same: target skills, situations observed, mastery level, decision. For apprentices, the framework varies by country. In France, for example, the apprenticeship supervisor (maître d'apprentissage) is tasked with contributing to the apprentice's acquisition of the skills required for the qualification, in liaison with the apprentice training centre (French Labour Code, article L6223-5). The form then complements the documents required by the school or training centre.
Looking to structure your workplace mentoring and better manage your teams' skills development? Daylindo digitises frontline skills tracking: mentors record their observations directly at the workstation, learning outcomes are tracked, and managers get a clear view of the skills mastered and where further support is needed.
