Your people attended the workshop. The feedback forms were glowing. Completion rates looked fine. And yet, three months later, almost nothing has changed about how work actually gets done. This is not a training delivery problem. It is a design problem. And the signs were there before a single session ran.
In 2025, Watershed and GP Strategies published their ninth annual survey on learning measurement. The opening finding carries a quiet indictment of the entire field: 98% of L&D professionals say they want to measure the business impact of learning. That figure has sat above 86% every year since the survey began in 2017. And yet, in the same 2025 report, only 27% of organisations set aside any specific budget for measurement activities — the lowest figure recorded since tracking began.[1]
That gap is not an oversight. It reflects a discipline that has, for structural reasons, learned to measure what is convenient rather than what is meaningful. Satisfaction scores, completion rates, and post-training quiz results are easy to collect and easy to report. What they are not, the evidence suggests, is reliable indicators of whether training changes how people perform at work.
The consequences show up in the LinkedIn Workplace Learning Report 2025, which surveyed 937 L&D and HR professionals alongside learner data from one billion LinkedIn members. The finding that stands out is not about AI or new delivery formats: it is that 49% of learning and talent development professionals now say their executives are concerned that employees do not have the right skills to execute business strategy.[2] These are organisations with active L&D programmes, training budgets, and learning platforms. The skills gap is not closing, and the people closest to L&D are the ones saying so.
This article does not argue that learning and development is failing as a function. Most learning practitioners are thoughtful, committed professionals who care about the people they serve. The problem is systemic. Organisations have built evaluation habits, reporting frameworks, and programme designs around the wrong indicators. The five signs below are not edge cases. They appear, in various combinations, in almost every L&D programme that struggles to demonstrate a link to business performance. Recognising them is the first step toward redesigning around what actually matters.
The post-session feedback form — rating content, facilitator, and logistics — is among the most persistent artefacts in corporate L&D. It survives not because it is useful, but because it is frictionless to collect, easy to aggregate, and straightforward to present upward. The problem is that it measures the wrong thing with remarkable precision.
A 2024 peer-reviewed study published in the journal Pharmacy tested the relationship between satisfaction during practical training and actual knowledge acquisition among students. The finding was unambiguous: despite recording significantly high satisfaction levels during training, there was no marked correlation between participant satisfaction and subsequent knowledge of the subject matter. The researchers concluded that satisfaction during training does not guarantee a thorough understanding, and that relying solely on satisfaction metrics is an insufficient basis for evaluating training programmes.[3]
This is consistent with the broader pattern in training evaluation research, which has repeatedly found that satisfaction at Level 1 of the Kirkpatrick model shows only weak to no correlation with learning at Level 2 and behavioural change at Level 3. What participants feel about a training session is shaped by a wide range of factors, including the personality of the facilitator, the quality of the catering, how much participants already knew about the topic, whether the timing was convenient, and whether the room temperature was comfortable. None of these factors have a direct relationship to whether the training produced a change in how someone performs at work.
The practical implication is significant. If your organisation's primary evidence that L&D is working is a satisfaction score above a threshold, you are not measuring performance impact. You are measuring the experience of attending training. The Watershed 2024 report adds an important structural dimension here: it found that L&D teams classified as "strategic partners" measure success by organisational impact, while those classified as "shared services" rely predominantly on satisfaction and content utilisation.[1] The measurement framework is not just a reporting choice. It reflects a deeper question about what the L&D function believes it is for.
Ask an L&D team how a programme came to exist, and the answer is often some version of: "The business said they needed training on X." The topic becomes the design brief. It should not be. Content-topic design asks what people should know about a subject. Performance-gap design asks a different question: what do people currently do, what do they need to do instead, and what is preventing that gap from closing? These two questions produce fundamentally different programmes.
The research on transfer of training is instructive here. A 2025 review published in the European Journal of Work and Organizational Psychology synthesised evidence on what makes learning transfer to the workplace. Three categories of factors consistently emerge: learner characteristics (including the degree to which they believe training is relevant to their actual work), training design features (especially the presence of practice opportunities and feedback that mirrors real work conditions), and the work environment after training (supervisor support, opportunity to apply, absence of situational barriers).[4]
Notice what is absent from that list: topic coverage. How comprehensive the curriculum was, how many modules the programme contained, or how well-structured the knowledge taxonomy is, none of these appear as reliable transfer drivers. Yet most L&D briefs are written almost entirely around content scope: how many sessions, which frameworks, how long each module runs. The design conversation begins and ends at the content level, never reaching the performance and context questions that determine whether training changes anything on the ground.
The result is a persistent and documented pattern. The LinkedIn Workplace Learning Report 2025 notes that 49% of learning leaders report executives are concerned that employees cannot execute business strategy, even in organisations with active training investments.[2] This is not evidence of a training volume problem. It is evidence that the training, however much of it has run, was not designed to close the specific performance gaps that strategy execution requires. A shorter, more contextualised programme anchored to a specific performance outcome will almost always outperform a comprehensive topic-rich programme that covers everything but is connected to nothing specific.
The World Economic Forum's Future of Jobs Report 2025 projects that 39% of existing skill sets will be transformed or made obsolete between 2025 and 2030, and that 59% of the global workforce will need reskilling to remain effective in emerging roles.[5] For L&D teams, the operative question is not whether the skills challenge is real. It is whether the training currently being designed is connected to the specific capability gaps that the organisation's strategy depends on closing. These are different questions, and most L&D programmes address the first without adequately addressing the second.
The manager problem is the transfer problem.
Research consistently identifies manager support as one of the strongest predictors of whether training transfers to the workplace. A 2025 review in the European Journal of Work and Organizational Psychology specifically identifies supervisor support, opportunity to apply knowledge, time resources, and feedback as the primary environmental factors that determine transfer success.[4] The mechanism is not complicated: if a manager never references the training, never asks how the learner is applying new skills, and never creates room to practise, the training sits in a separate context from work.
The LinkedIn Workplace Learning Report 2025 adds a concrete data point to this picture: only 15% of employees say their manager helped them build a career plan in the past six months, a decline of five percentage points from the year before.[2] If managers are not engaged in even relatively light development conversations, they are almost certainly not actively reinforcing learning transfer either. L&D teams that design programmes without a manager activation strategy are, in effect, designing for an environment that does not exist in practice.
There is a moment in almost every L&D programme that rarely gets examined because it is the moment the team's attention naturally shifts elsewhere. It is the moment the last session ends, the facilitator packs up, and participants return to their desks. From a behaviour change perspective, this is not the finish line. It is closer to the starting line.
The evidence on learning consolidation is unambiguous on one point: what happens after training matters as much as what happens during it. The opportunity to practise newly learned behaviours in real work conditions is a necessary condition for transfer to take hold.[4] Without a manager who creates space for new approaches, a work environment that accommodates the temporary adjustment that comes with applying new skills, or any follow-up mechanism that brings the learning back into contact with real work, the programme fades. This is not a weakness of the training. It is a predictable outcome of how learning consolidates into changed behaviour.
The TalentLMS 2026 L&D Report, drawing on survey data collected through 2025, captures a related and troubling pattern. Multitasking during training has reached its highest level in three years, with 70% of employees admitting to checking email or working on other tasks during training sessions in 2025, up from 58% in 2024.[6] When participants are not fully present during training, the baseline of what gets learned is already compromised. Designing strong post-training reinforcement for a programme where cognitive engagement was partial during delivery is working against compounding disadvantages.
The signal to look for in your own organisation is structural: is there any planned activity in the weeks after a programme ends? Are participants asked to demonstrate application of what they learned? Are managers briefed on what was covered and asked to create opportunities to use it? Is there any mechanism at all that reconnects learning to live work? If the answer is no, the programme ends at precisely the wrong moment.
There is a distinction in how L&D teams position themselves relative to the business that has significant consequences for impact, and it rarely gets named directly. Some L&D functions operate as strategic partners: they are involved in business planning conversations, they interrogate the capability gaps behind performance problems, and they co-design interventions with the business. Others operate as shared services: they fulfil requests, build what they are asked to build, and deliver to the specifications they are given. These two models produce measurably different outcomes.
The Watershed and GP Strategies measurement report consistently tracks this distinction. Strategic partner L&D teams measure success by organisational impact, align their metrics with business KPIs, and are more closely integrated with the rest of the organisation. Shared services teams measure success primarily by satisfaction and content utilisation. The split is not just a reporting difference. It reflects a fundamentally different theory of what L&D exists to do.[1]
The shared services model has a seductive logic. It is responsive. It avoids the political friction of pushing back on a stakeholder's framing. It keeps the L&D team in a position of neutral provider rather than opinionated partner. And in the short term, it produces satisfied stakeholders: the team delivered what was asked, on time and to brief.
The problem is that stakeholders frequently do not know what they need. A sales director who requests communication training for an underperforming team has identified a symptom and proposed a solution. They have not necessarily identified the cause. The cause might be unclear product knowledge, misaligned incentives, or weak pipeline management. Communication training will not address any of these. An L&D function that builds the programme without interrogating the diagnosis is delivering efficiently toward the wrong outcome.
PwC's 2025 Global Workforce Hopes and Fears Survey, drawing on responses from nearly 50,000 workers across 48 countries, found that upskilling efforts remain structurally uneven at most organisations: just 51% of non-managers report having access to the learning and development resources they need, compared to 66% of managers and 72% of senior executives.[8] That gradient is itself a diagnostic signal. When learning investment disproportionately reaches those already closest to decision-making, the L&D function is likely responding to the loudest voices rather than the deepest performance gaps. The employees doing the most operational work are the least served by the very function designed to develop them.
A performance-diagnostic L&D conversation starts with different questions: What is the specific gap between current and required performance? How do we know this is a knowledge or skill gap rather than a process, incentive, or resource problem? Who exactly is underperforming, in what situations, and what do high performers do differently? These questions often reveal that the requested training is either the wrong solution entirely, or one component of a multi-part intervention. That discovery is what separates L&D that changes performance from L&D that fills a training calendar.
The dominant model of corporate training, assembling people in a room or on a platform for a defined period to receive structured content, was inherited from an era when learning and work were separated by practical necessity. That separation has hardened into a design assumption. And it is one of the most expensive assumptions in the modern organisation.
The World of Learning 2024 report, a practitioner-focused research publication, identifies the central issue directly: effective learning requires moving beyond course-centric methods toward integrating learning and performance support into work itself.[9] Integrating into work is the operative phrase. Not running alongside work. Not pausing work to attend training. Connecting learning to the conditions of real work, so that the moment of need and the moment of learning are as close together as possible.
When learning is extracted from work context, several things happen. The practice scenarios in the training are necessarily artificial. The feedback participants receive during training is not the same feedback they will receive when performing the skill at work. The cognitive and social conditions of the training environment differ from the conditions where the learning must eventually apply. Transfer research is consistent on this point: the more different the training environment is from the performance environment, the harder it is for learning to transfer.[4]
McKinsey's research on capability building identifies the same structural issue from an organisational strategy perspective. The most effective organisations build learning that is skill-specific, contextually relevant, and connected to real career pathways, rather than structured around generic topic catalogues and cohort schedules.[10] The implication is that formal training, isolated from the specific work context of the learner, has a limited and often diminishing role in building the capability the organisation actually needs.
The sign is visible in how your learning calendar is constructed. If training events and business cycles are planned independently of one another, if leadership programme cohorts run in October regardless of what the business is doing in October, if module sequences are driven by curriculum logic rather than by when learners have live opportunities to apply each topic, the programme is designed around the convenience of delivery rather than the conditions for learning transfer. These are different designs, and they produce different results.
Training timelines are designed around when people have live opportunities to practice, not around facilitator availability. Job aids and reference tools exist at the point of performance, not buried in an LMS portal that participants stop visiting after the programme ends.
Managers are briefed before programmes run, so they can create application opportunities immediately after. Follow-up check-ins are scheduled as part of the programme design, not added as optional extras. Completion is defined by demonstrated application, not attendance.
When the gap between learning moment and performance moment shrinks, the probability that learning changes behaviour increases. This is a design principle, not a delivery preference.
Training calendars are built around topic availability rather than performance cycles. Programmes assume learners arrive without specific context and will return to identical conditions after completing the course. Content is generic enough to serve multiple audiences but specific enough to serve none of them particularly well.
Completion certificates are issued based on attendance and quiz scores, with no assessment of whether the learner can apply the knowledge in a real work context. LMS data shows completion rates but cannot speak to any downstream performance metric.
The most expensive training calendar is one that runs consistently, is consistently well-attended, and consistently fails to move the performance numbers that business strategy depends on.
The five signs above are not independent failures. They are symptoms of the same root condition: L&D designed around the activity of training rather than the outcome of performance. Addressing one without addressing the others produces partial progress at best.
The underlying question, the one that exposes the full scope of the challenge, is this: if you had to demonstrate, with specific evidence, that your L&D investment changed how people in your organisation perform at work, could you do it? Not that they attended programmes. Not that they completed modules. Not that they rated the experience highly. That they do something demonstrably differently at work because of what the L&D function did.
For most organisations, the honest answer remains no. That is not a reflection of the effort or care that learning professionals bring to their roles. It is a reflection of the system those professionals are operating within: evaluation frameworks built around the wrong indicators, design processes that begin at content rather than performance, and a relationship with the business that remains transactional rather than diagnostic.
The stakes are not abstract. The WEF Future of Jobs Report 2025 projects that 59% of the global workforce will need reskilling by 2030.[5] PwC's 2025 Global Workforce Hopes and Fears Survey found that just 51% of non-managers feel they have the L&D resources they need, at a moment when the cost of that gap is accelerating.[8] Against that backdrop, an L&D function that cannot demonstrate it is closing the capability gaps that matter is not a quiet cost centre. It is a strategic risk.
The rewiring required is not primarily technological. It is conceptual. It means redefining what L&D is for: not the management of training activities, but the engineering of performance conditions. It means building the measurement infrastructure before the programme launches rather than retrofitting a business case after the feedback forms come in. It means earning the right to be in the room where performance problems are diagnosed, not only in the room where content solutions are specified and delivered.
None of this is simple. But the first step is seeing the signs clearly, and being willing to name what they indicate.
If you recognise any of these signs in your current L&D investment, the conversation worth having is about programme architecture, not content scope. Baugment designs and develops competency-based learning solutions built around specific performance outcomes, with measurement aligned from the start. Get in touch.
