Measuring impact

How to measure the ROI of training (and prove it to your board)

AH

Ash Heeramun

Founder, Apex Minds · 10 Nov 2026 · 7 min read

67%

of leaders recognise L&D's impact, down from 81% in 2021

Measuring impact

In short

Training ROI (%) = (monetary benefit of the training − total cost of the training) ÷ total cost × 100. The formula is the easy part. The hard part is the benefit figure, and that only becomes credible if you set a baseline before the training starts, choose one measure that moves within weeks and one business outcome that moves within months, and report the change at team level rather than relying on course feedback.

The training ROI formula, explained

Most guides stop at the formula: ROI (%) = (Benefits − Costs) ÷ Costs × 100.

  • Costs are everything the programme takes: provider fees, licences, internal design time, and the hours employees spend away from their work.
  • Benefits are the money value of what changed: more output, fewer errors, lower staff turnover, less time lost to sickness, faster time to competence for new starters.

Two well-known frameworks sit behind most ROI work. The Kirkpatrick model evaluates training at four levels: reaction (did people like it?), learning (did they learn it?), behaviour (do they do things differently?) and results (did the business improve?). The Phillips ROI methodology adds a fifth level that turns those results into money and compares them with cost.

Most organisations measure the first two levels well and the last three poorly.

Why most L&D teams struggle to prove impact

Almost half

of organisations lack a systematic evaluation and measurement approach

CIPD, 2023

67%

of leaders recognise L&D's impact on priorities, down from 81% in 2021

CIPD

30%

of organisations believe their capability programmes deliver measurable impact

McKinsey, via CIPD

CIPD's Learning at Work 2023 survey found that almost half of organisations lack a systematic evaluation and measurement approach, and fewer than one in ten L&D practitioners strongly agree they make evidence-informed decisions to improve performance.

The same CIPD research recorded a fall in the proportion of leaders recognising L&D's impact on organisational priorities, from 81% in 2021 to 67% in 2023. McKinsey research cited by CIPD found that 78% of organisations see capability building as vital to growth, but only 30% believe their programmes deliver measurable impact.

LinkedIn's 2025 Workplace Learning Report found that 49% of L&D and talent professionals say their executives are concerned employees don't have the skills to execute the business strategy, and that engagement and retention are currently the most common ways to measure the business impact of career development.

In other words, the board is asking whether learning moves the business, and most L&D teams can only answer with attendance and satisfaction scores.

Step 1: Start with the business problem, not the course

Write down the one business outcome the training is meant to change, in plain numbers: "reduce first-year sales team turnover", "cut rework in operations", "shorten time to target for new account managers". If you can't name the outcome, you can't calculate a return on it.

Step 2: Set a baseline before anything starts

Record where that outcome sits today, and for which teams. Without a "before" figure, every "after" figure is a guess. If you can, keep a comparison group: a team that starts the programme a quarter later gives you a fair benchmark.

Step 3: Choose one leading measure and one lagging measure

Business outcomes such as turnover or revenue per head take months to move. Pair each one with a leading measure that moves within weeks: focus, confidence, energy, stress levels or a skills check. Leading measures tell you early whether the programme is working, so you can adjust before the money is spent.

Step 4: Count the full cost

Include the time employees spend away from their work, not just the invoice. Finance teams notice when it's missing.

Step 5: Report at team level at 30, 90 and 180 days

Report the change against your baseline for each department, not individual scores. Team-level results protect privacy and show where the programme is working and where it isn't.

A worked example (illustrative figures only)

A firm spends £18,000 on a six-month programme for a 60-person operations department, including employee time. Before the programme, rework cost the department about £9,000 a month. Six months later it costs £6,000 a month, while a comparison department shows no change. The saving attributable to the programme is £3,000 a month, or £36,000 a year.

ROI = (£36,000 − £18,000) ÷ £18,000 × 100 = 100%.

These numbers are invented to show the calculation; they are not a benchmark.

The gap: you only measure the people you train

Coaching and leadership programmes go to top performers. Employee assistance programmes (EAPs) pick people up after something has gone wrong. Everyone in between, the majority who carry day-to-day delivery, gets little built for them and even less that is measured.

That is why we talk about raising the performance floor, not just the ceiling. If you only measure the people you coach, you miss where much of the return can come from: the middle of the organisation.

Apex Minds is a cognitive performance platform built around three steps: Train, Measure, Prove.

  • Train: employees use short, personalised audio programmes during the working day to reset, refocus and recover.
  • Measure: the Apex Minds Index reports Power, Wellness and Performance scores for each department, RAG-rated and anonymised. It reflects employees who use the platform; it does not identify individuals.
  • Prove: the employer dashboard brings together engagement, sentiment and trend reporting, giving you the leading measure in step 3 and a team-level report for step 5.

Frequently asked questions

What is a good ROI for training?

There is no universal benchmark: it depends on the cost of the problem the training is solving. A useful rule is to agree the target with your finance director before the programme starts, so the result is judged against a figure everyone accepted in advance.

What is the difference between the Kirkpatrick model and Phillips ROI?

Kirkpatrick evaluates training at four levels: reaction, learning, behaviour and results. Phillips adds a fifth level that converts results into money and compares them with the cost, giving a percentage ROI.

How long does it take to see a return on training?

Leading measures such as focus or confidence can shift within weeks. Business outcomes such as turnover, error rates or revenue per head usually need three to six months, which is why a baseline and a 30/90/180-day reporting rhythm matter.

Can you measure the ROI of soft-skills or resilience training?

Yes, if you link it to a business outcome and pair it with a leading measure. Resilience training that is only measured by a feedback form can't show a return; resilience training measured against team-level focus, energy and absence data can.

How do I measure training impact without identifying individuals?

Report at department or team level and compare against a baseline. This is how the Apex Minds Index works: anonymised, department-level scores rather than individual results.

Sources

Raise the performance floor, not just the ceiling

Book a demo or start a pilot.

Book a Demo