Going Beyond Completion Rates in Learning

In 2026 most organisations can tell you how many employees enrolled in a course, how many completed it, what assessment scores they achieved, and whether their learners were satisfied. L&D has seemingly never had more data at its fingertips.

But the harder question to answer is what happened next: Did employees become more capable? Did they apply what they learned? Did performance improve? And, ultimately, did the organisation move closer to the outcome the learning initiative was intended to support?

That gap between measuring learning activity and demonstrating business value is becoming increasingly difficult to ignore. Gartner's guide has highlighted growing pressure on L&D leaders to communicate their function’s value in terms that connect with wider enterprise priorities, while Scheer IMC’s own State of Learning Technologies 2026 research found a similar challenge: 44% of surveyed enterprise L&D decision-makers reported difficulties linking their learning activities to wider business impact.

The answer is not necessarily to collect more data, but to become clearer about what should be measured in the first place.

Traditional learning metrics remain useful as low completion rates may indicate that employees cannot find time for training. Equally, poor assessment results could reveal that the content is ineffective or too difficult.

But problems arise when these measures are treated as markers of progress in their own right...

Completing a leadership programme, for example, tells us that someone participated, and if they passed its final assessment that may tell us that they understood the concepts. But neither metric necessarily tells us whether their behaviour as a manager changed six weeks later.

The same applies to sales training, onboarding, or compliance. The closer a measure sits to the learning activity itself, the less it tells us about what subsequently changed in the workplace. This means for senior stakeholders asking what an L&D investment achieved, “92% completion” is therefore rarely the end of the story.

Recognising the limits of learning activity metrics can lead organisations to the opposite extreme: trying to attach a monetary return to every learning initiative.

Financial ROI can be extremely useful where there is a clear relationship between an intervention and an economic outcome. But learning usually exists within a much larger system.

Sales performance, for example, might be influenced by training, but also by market conditions, pricing, product changes, sales management, and lead quality. Employee retention can be affected by development opportunities, but also by pay, leadership, workload, and career prospects.

Trying to attribute all improvement to learning can therefore create a false sense of precision.

It can be more important for enterprises to ask themselves: What would credible evidence of success look like for this particular business problem?

But the answer will differ from one learning initiative to another:

For onboarding, reducing time to proficiency may be more meaningful than calculating a theoretical monetary return.
For compliance training, risk reduction, fewer incidents or fewer audit findings may be more relevant.
For reskilling, validated proficiency and successful deployment into new roles may provide stronger evidence.
For sales enablement, the connection to business performance may be closer, allowing learning data to be considered alongside metrics such as conversion, deal quality or win rate.

Impact is often broader than a single formula.

One way to think about learning impact is as a progression:

ACTIVITY → LEARNING → CAPABILITY → APPLICATION → PERFORMANCE → BUSINESS OUTCOME

Each stage answers a different question.

Activity: Did people participate?
Learning: Did they acquire the intended knowledge or understanding?
Capability: Can they perform the required skill?
Application: Are they using it in their work?
Performance: Has the relevant aspect of job performance changed?
Business outcome: Has this contributed to the wider organisational objective?

Not every programme needs to be measured at every stage, and nor does every initiative need to reach the final step. The value of the model is that it forces L&D teams to ask where their current evidence actually stops. An organisation may well discover, for example, that it measures activity and learning extremely well but has almost no visibility into capability or application, and that makes it difficult to explain why training should influence a business outcome several steps further down the chain.

This is where learning analytics can become useful – not simply for producing more dashboards, but for bringing together different sources of proof that help L&D understand what is happening beyond participation alone.

Learn more about how Scheer IMC can support with establishing learning analytics for more measurable ROI.

For many organisations, the most useful measures may sit somewhere between completion rates and business KPIs, and it could be capability and proficiency that can provide this middle layer.

The World Economic Forum's Future of Jobs Report 2025 found that skills gaps remain one of the biggest barriers to business transformation, while 85% of surveyed employers planned to prioritise workforce upskilling. If organisations are investing in learning primarily to build capabilities, then evidence of capability development becomes important in its own right.

This may include:

demonstrated proficiency in an assessment or realistic scenario
improved performance in simulations
manager or peer observation
reduced time to competence
successful application of a new skill in the workflow
internal mobility into roles requiring newly developed capabilities

These indicators do not replace business outcomes, but instead help explain how learning contributes to them. For many learning initiatives, that provides a more credible narrative than jumping directly from “course completed” to “business performance improved”.

One of the biggest measurement mistakes happens before any data is collected. If success is only discussed once the programme has finished, L&D teams often find themselves searching retrospectively for metrics that might demonstrate value. CIPD’s guidance on learning evaluation takes the opposite approach: start with the performance gap.

That means asking a different set of questions at the beginning:

What business or performance problem are we trying to influence?
What should employees be able to do differently?
Which capability or behaviour needs to change?
What evidence would show us that this happened?
Where does that data already exist?

Only then should the learning intervention and its measurement approach be designed. This also changes the conversation with business stakeholders such that, instead of asking them to approve “a training programme”, L&D can agree on the performance problem, the intended change, and the evidence that would indicate progress. Measurement then becomes part of the learning strategy rather than an exercise added at the end.

This challenge becomes even more important as AI changes how learning content is produced. Tools can already accelerate activities such as content creation, localisation and translation. As these capabilities become more widely adopted, businesses may be able to produce and update learning materials significantly faster.

But a higher volume of content does not automatically mean greater organisational impact. If anything, easier content production makes prioritisation more important.

L&D teams should stop asking themselves: "how much learning did we create?" and instead pose the question: "did we develop the capabilities the organisation actually needs?"

That shifts attention from production volume towards relevance, application, and measurable change.

Ultimately, there remains no universal metric that "proves" learning has created value. A completion rate can be important, as can a skills assessment. Equally important may be metrics such as time to proficiency, operational performance, or financial ROI.

But the broader point remains that the right measure depends on the problem the organisation was trying to solve. For L&D and HR leaders, the takeaway is therefore not to build increasingly larger dashboards, but rather place the focus on creating a clearer line between the business need, the capability being developed, and the evidence used to judge success.

The most useful final outcome is likely to be more than just whether employees learned something. It is also what they can do differently as a result – and why that matters to the organisation.

Going Beyond Completion Rates in L&D

Why "traditional" learning success metrics don't tell the full story

Completion rates tell us something – just not everything

But financial ROI is not always the answer either

Build an evidence chain from learning to performance

Capability is often the missing middle

Remember that learning does not operate in isolation

AI makes output a weaker measure of value

From measuring learning to understanding what changed

The L&D Impact Gap: What 420 Decision-Makers Told Us

From Business Goal to Learning Metric: A Practical Impact Measurement Guide