Measuring L&D impact UK is one of the most persistent frustrations in the profession. L&D teams invest significant time and budget in training, yet when a senior stakeholder asks "what difference did it make?", the honest answer is often: we do not have the data to say.
This is not a new problem. But the CIPD's Learning and Skills at Work research, produced in partnership with Accenture, confirms it remains a live one. The report examines current practices and trends across UK organisations and finds that linking learning activity to business outcomes continues to be a significant challenge for most L&D functions.
The question worth asking is not "why can't we measure impact?" but "why was the learning never built to produce evidence in the first place?"
The barriers CIPD identifies are structural, not technical
The CIPD press release accompanying its Learning and Skills at Work findings is direct: L&D professionals face "significant challenges" in delivering on their priorities, and the CIPD calls on practitioners to "leverage data and insights" to bridge the skills gap. That framing matters. It acknowledges that the data problem is real, and that it sits at the heart of the profession's credibility with business leaders.
The barriers are structural. L&D teams are often under-resourced, evaluation sits low on the organisational priority list, and the design and delivery roles that could build measurement into programmes from the start are stretched thin. When capacity is tight, evaluation is the first thing cut - and it is usually cut before the project even begins, not after.
This creates a compounding problem. If evaluation is not scoped at the commissioning stage, the learning that gets built will not generate the data needed to report on. No amount of post-launch survey-chasing or spreadsheet-wrangling will recover what was never captured.
Evaluation is not a phase - it is a design decision
The CIPD's factsheet on learning evaluation, impact and transfer sets out approaches for measuring whether learning engages employees, transfers to the workplace and delivers organisational impact. The frameworks it describes - including well-established multi-level models - all share a common requirement: you need to know what you are measuring before you build the learning, not after you have delivered it.
This is the distinction that most organisations miss. Evaluation is treated as a reporting task that happens at the end of a project. In practice, it is a design decision that shapes every element of how a course is built, what interactions it includes, how it is tracked and what data the LMS can surface.
Consider what this means in practice:
- Completion and pass rates require SCORM or xAPI tracking to be configured correctly at build stage, with the LMS set up to receive and store that data.
- Knowledge retention requires assessment questions designed to test application, not just recall, with results tied to learner records.
- Behaviour change requires a baseline measure before training and a follow-up mechanism after it - neither of which can be retrofitted once a course is live.
- Business outcome linkage requires the learning objectives to be mapped to measurable performance indicators from the outset, not described in vague terms like "raise awareness."
None of these are difficult to build in. All of them are very difficult to add afterwards.
What "designed for evidence" looks like in practice
When a learning programme is commissioned with evaluation in mind, the brief looks different. Objectives are written in terms of observable behaviour change. Assessment design is discussed alongside content design. SCORM or xAPI is specified as a delivery requirement, not an afterthought. The LMS reporting structure is agreed before a single slide is built.
This is not about adding complexity or cost. It is about making deliberate choices early that prevent expensive retrofitting later. A well-structured xAPI implementation, for example, can capture not just completion but time-on-task, branching decisions, assessment attempts and scenario outcomes - all of which give L&D leaders something meaningful to report to the business.
The difference between a course that produces evidence and one that does not is rarely the subject matter or the budget. It is whether the people commissioning and building it asked the right questions at the start.
The organisational priority problem
One of the more uncomfortable findings in the CIPD's research is that evaluation sits low on organisational priority lists. This is partly a resourcing issue, but it is also a cultural one. If senior leaders do not ask for evidence of learning impact, L&D teams are unlikely to invest in producing it. And if L&D teams do not produce it, senior leaders have no basis on which to increase investment.
Breaking this cycle requires L&D leaders to make the case for evaluation at the commissioning stage - not as an optional extra, but as a condition of building anything at all. That means being explicit with stakeholders about what data the programme will and will not generate, and what decisions that data can support.
It also means being honest about the limits of what can be measured. Not every learning intervention will produce a clean ROI calculation. But completion rates, assessment scores, learner satisfaction and manager-reported behaviour change are all legitimate forms of evidence that can be reported with confidence - provided the learning was built to capture them.
Alistair's take
The CIPD's research reflects something we see consistently in practice. Organisations commission learning, deliver it, and then find themselves unable to answer basic questions about whether it worked. The problem is almost never the LMS or the reporting tool. It is that the learning was not built with reporting in mind.
At Neon, we have supported more than 40 clients across a range of sectors, and the projects that generate the most useful impact data are the ones where evaluation was part of the brief from day one. That means agreeing on what success looks like before any content is written, specifying SCORM or xAPI tracking as a delivery requirement, and designing assessments that test the behaviours the organisation actually needs to change.
The CIPD is right that L&D professionals need to leverage data and insights. But that leverage only exists if the data was designed in. Retrofitting evaluation onto content that was never built to produce it is one of the most common and most avoidable problems in the profession.
Build learning that produces evidence from day one
Neon Learning designs and builds custom eLearning with SCORM and xAPI tracking, structured assessment and LMS reporting built in from the commissioning stage - so you have something meaningful to report when it matters.