October is the month when most executive teams begin defining the following year’s training plan. It is also when they make the decisions that determine whether that plan will be a serious response to the business’s real needs or an administrative exercise repeated with minor variations on the previous year. The difference between these outcomes lies neither in the budget nor in the chosen catalogue. It lies in the methodology used to design the plan. And for the first time in the history of corporate training, that methodology can draw on artificial intelligence to substantially improve diagnostic accuracy, dramatically shorten preparation time and produce a truly individualised plan across the entire workforce.
At CAE, we have spent more than forty-five years helping large companies, consultancies and universities build their training programmes. Over the last twenty-four months, we have seen how the emergence of artificial intelligence applied to talent analysis has radically transformed what is possible when designing an annual training plan. This article sets out a structured eight-step methodology for HR managers, talent leaders and executive teams seeking to build their 2027 training plan with sound judgement, drawing on the capabilities AI offers today while avoiding the common mistakes of traditional design.
Why the annual training plan remains the executive team’s most underestimated decision
Before discussing methodology, it is worth acknowledging a paradox in the sector that is rarely articulated. The annual training plan directly shapes the organisation’s ability to execute its strategy over the following twelve months. It determines which skills the workforce will have available, which projects it can confidently undertake, what pace of growth it can sustain and which operational risks it can avoid. Yet many executive teams apply less methodological rigour to this decision than they do to infrastructure, technology or commercial expansion.
This mismatch between importance and rigour has a historical explanation: for decades, there were no tools that allowed a training plan to be designed with the same analytical precision as other strategic decisions. Training needs analysis was, almost by definition, an exercise in expert estimation with wide margins of error. But that explanation no longer holds. The tools exist, they are mature and they are accessible. What is often missing is the decision to use them.
The 8 steps of a rigorous methodology for designing the 2027 training plan
In our experience, the following eight steps form a training plan design process that produces plans genuinely aligned with strategy and capable of delivering measurable impact. Each step draws on specific capabilities that AI offers today.
Step 1: explicitly define next year’s strategic objectives
The training plan starts with strategy, not with the catalogue. Before deciding what training to deliver, the organisation must precisely identify its strategic objectives for the following year, the projects that will require significant resources, the transformations it will undertake and the markets it will enter. This information must come from the executive team and the strategy team, rather than the training department, because objectives determine the skills needed, not the other way round.
Step 2: translate strategy into a map of critical skills
Once the objectives have been defined, they must be translated into the specific skills the workforce needs to execute them. This is probably the most delicate step in the process because it requires both business judgement and educational judgement. An objective such as expanding into Latin American markets translates into very specific skills in international sales, cross-cultural management, local regulatory compliance or coordination with remote teams, with expected proficiency levels for each role.
Step 3: automatically diagnose current skills gaps
This is where artificial intelligence delivers the first qualitative leap. Instead of manually gathering assessments and spreadsheets from the previous year, an AI engine such as Imagine automatically analyses each employee’s actual skills, cross-references that information with job requirements and the critical skills defined in the previous step, and produces a detailed skills gap map covering the entire workforce. What once took months can be completed in hours, with a level of precision that manual methods cannot achieve.
Step 4: strategically prioritise the gaps identified
Not all identified gaps deserve the same investment. Prioritisation should combine three criteria: each gap’s impact on strategic objectives, how urgently it must be closed before it affects the business, and the relative cost of closing it. As reflected in Deloitte’s annual Global Human Capital Trends reports, this prioritisation separates training plans that deliver real value from those that simply distribute the budget equally.
Step 5: automatically generate personalised learning pathways
Once the gaps have been prioritised, the AI system automatically generates a specific learning pathway for each employee, selecting the courses from the available catalogue that address each gap, with the most appropriate duration, sequence and delivery mode. This step, which traditionally required painstaking manual work for each job profile, becomes fully automated. The training team reviews and validates the proposals, applying its judgement to special cases, while most of the repetitive work disappears.
Step 6: integrate budgeting with available funding
Each pathway has an associated cost, and their combined cost forms the year’s total training budget. At this stage, the budget should be cross-referenced with available funding sources: FUNDAE training credit, sector-specific grants, business budgets for strategic programmes and the general HR budget. A well-designed plan makes the most of subsidised funding without allowing that optimisation to undermine the plan’s strategic coherence.
Step 7: define the monitoring dashboard
Before launching the plan, the organisation must define how its delivery and impact will be measured. At a minimum, the metrics should include activity indicators (course completion and time spent), quality indicators (satisfaction and assessment results), learning transfer indicators (application on the job measured through digital activity in corporate tools) and business impact indicators (the operational KPIs the training was intended to improve). This dashboard must be defined before implementation, rather than designed afterwards when it is too late to establish a baseline.
Step 8: establish governance and quarterly reviews
The annual training plan is not a closed document to be implemented in a straight line. It is a living framework that should be reviewed quarterly to adjust priorities as the business evolves, incorporate new critical skills as they emerge and reallocate budgets between programmes if some progress faster than expected. Shared governance between HR and business leaders ensures the plan remains aligned with operational reality throughout the year.
How AI transforms each of these steps
Taken together, these eight steps make it clear that artificial intelligence does not replace strategy or human judgement in plan design. It replaces the mechanical, repetitive work that consumes the training team’s time without adding value. AI does three specific things that substantially improve the resulting plan.
The first is analysis at scale. A human team can accurately diagnose skills gaps for fifty key employees. AI can diagnose them for ten thousand employees with the same precision. This eliminates the historical tension between depth of analysis and workforce size that has limited training plans for decades.
The second is continuous updating. A manually designed plan captures the diagnosis at the moment it was prepared and quickly becomes outdated. An AI-generated plan is continuously updated as employees’ skills, job requirements or strategic priorities change. As documented by publications such as Harvard Business Review, this capacity for continuous updating is probably the feature that most clearly distinguishes the emerging model from the traditional one.
The third is genuine personalisation. AI enables every employee to receive a pathway specifically tailored to their role, level, particular gaps and professional goals—something that manual methods made feasible only for senior executives with substantial individual budgets. Democratising personalisation is the change with the greatest transformative potential in the sector.
The four most common mistakes when applying this methodology for the first time
It is worth identifying the mistakes we most often see in organisations using this methodology to design their training plan for the first time.
The first is starting with the catalogue rather than the strategy. When an organisation begins by asking which courses are available on the market, the resulting plan is constrained by supply rather than driven by need. The correct sequence is always strategy → skills → gaps → pathways, not the reverse.
The second is failing to involve the executive team in the initial steps. When strategic objectives and critical skills are defined by HR alone, the plan loses alignment with business reality. Shared responsibility with the executive team for the first two steps is a non-negotiable condition of a rigorous plan.
The third is delegating pathway validation entirely to AI without human review. Modern systems produce very robust recommendations, but there are always special cases—critical roles, role transitions or employees’ personal circumstances—that require specific human judgement. Human governance is not optional; it is part of the method.
The fourth is failing to establish a baseline before starting. Without a prior measurement of the impact indicators the training aims to improve, there will be no way to demonstrate the plan’s return at the end of the year. Defining the baseline must be part of the design process itself, rather than an afterthought.
A realistic schedule for designing the 2027 plan in the fourth quarter of 2026
For training managers beginning to design their 2027 plan now, the following is a realistic three-month schedule.
During the second half of October, the organisation should finish defining its strategic objectives with the executive team and translating them into a map of critical skills. This stage involves the greatest number of meetings and the most coordination with other departments.
During the first half of November, automated gap diagnosis and strategic prioritisation should take place. If the organisation already has an active AI engine, this stage can be completed in a few days. If it is the first time, the initial configuration period must also be allowed for.
During the second half of November and the first half of December, personalised pathways should be generated, budgets integrated and the dashboard defined. This is the stage with the highest concentration of detailed work.
During the second half of December, the executive team should formally approve the plan, managers should receive the internal communication and preparations should be completed for a January launch. A plan approved in December can get off to a clean start in the first quarter; a plan approved in January will inevitably lose that quarter.
How Imagine fits into this methodology
At CAE, we designed Imagine specifically as the operational enabler that makes this methodology feasible at genuine corporate scale. Imagine automates steps 3 and 5 (gap diagnosis and pathway generation), provides technical support for steps 4 and 6 (prioritisation and budgeting), and automatically generates the data that feeds step 7 (the dashboard). The HR team retains full strategic responsibility for steps 1, 2 and 8 (defining objectives, translating them into critical skills and ongoing governance), precisely where human judgement is irreplaceable.
Combining Imagine with the LearningHub CAE catalogue, which provides more than eight thousand five hundred continuously updated courses for the engine to select from, completes the equation. The result is a 2027 training plan designed with methodological rigour, individualised for every employee, aligned with strategy and capable of delivering measurable impact from the first quarter.
Frequently asked questions
Is this methodology feasible if it is the organisation’s first experience of AI in training?
Yes, provided the first year is approached as an exercise in organisational learning. Technical activation can be completed in weeks, but embedding the new method culturally takes between three and six months. Organisations that accept this learning curve achieve sustainable results from the second year onwards.
Can this methodology be combined with the use of FUNDAE training credit?
Absolutely; in fact, it is an especially good fit. Automatically generating pathways helps optimise the allocation of subsidised training credit to the categories with the greatest strategic impact, avoiding the scattered approach typical of the traditional model. The system automatically maintains all the SCORM or xAPI tracking required for the subsidy.
Does AI replace the HR team in designing the plan?
No. It frees HR from repetitive work so the team can focus on strategic decisions where its judgement is irreplaceable. The team spends more time defining objectives, prioritising critical skills and supporting managers, and less time on the mechanics of assigning courses and building spreadsheet matrices.
How long does the full process take, from the first step to plan approval?
With this methodology, it takes between ten and twelve weeks when the organisation already has an active AI engine. In the first year of implementation, allow an additional four to six weeks for the initial setup. Compared with the four or five months required by the traditional method, the time saving is significant.
What happens if the organisation’s strategy changes halfway through the year?
One of the distinguishing features of an AI-designed plan is that it can be updated dynamically when priorities change. The engine recalculates the relevant gaps, adjusts priorities and generates updated pathways for the employees affected. This turns the plan from a closed document into a living organisational capability.
Conclusion: for many organisations, the 2027 plan will probably be their first designed with AI
The transition from manually designing the training plan to AI-assisted design is probably the methodological change with the greatest potential impact on the HR function in the coming years. And for many Spanish organisations, 2027 will be the first training plan designed using this new methodology.
Organisations that approach this transition with methodological rigour will gain three benefits simultaneously: training plans genuinely aligned with strategy, delivery with measurable impact on business indicators, and HR teams repositioned as strategic partners to the executive team. Those that continue applying the traditional method while competitors adopt the new approach will operate at a growing structural disadvantage.
After forty-five years, we at CAE remain convinced that well-designed corporate training is one of the most profitable investments an organisation can make. The emergence of artificial intelligence strengthens that conviction, because for the first time it is possible to apply full methodological rigour to decisions that relied on expert approximation for decades.
Would you like to find out how to apply this methodology to your 2027 training plan? Contact our team for an initial conversation about your specific context and the practical steps involved in the transition.