← All articles

“Training is expensive.” You’ve probably heard that line. But what does training that never happens cost? Without numbers, training is often the first expense to be cut. With four simple numbers, you can show what it really brings in.

Key takeaways

  • Track four numbers before and after training: time to autonomy, errors and scrap, incidents, and departures in the first 90 days.
  • The return on investment calculation is simple: gains minus costs, divided by costs.
  • Take a baseline before the training. Without it, you can’t prove the effect.

In this article

  1. Why measure
  2. The 4 numbers to track
  3. Calculating return on investment
  4. Set up measurement in 5 steps
  5. Traps to avoid
  6. The cost of not training

Why measure

When production is under pressure, training often comes last. It’s seen as an expense, not an investment. The only way to change that view is to show results in numbers.

Good news: you don’t need a complicated system. You already have almost all the data. You just need to look at it before and after.

The 4 numbers to track

⏱️

Time to autonomy

How many weeks does a new employee need to work alone at the right pace?

How: record the hire date and the date they become autonomous in your matrix.

🔧

Errors and scrap

How many rejected parts, reworks or errors on the targeted job?

How: use your quality reports, by month and by job.

🩹

Incidents

How many first-aid cases, near misses and accidents on that job?

How: use your accident and incident log.

🚪

Early departures

How many new hires leave within the first 90 days?

How: compare hire and departure dates.

Pick one or two numbers linked to the training you want to evaluate. Training on one job mainly targets errors and autonomy. An onboarding program mainly targets early departures.

Calculating return on investment

The formula

(Gains − Costs) ÷ Costs × 100 = ROI in %

Costs are building or buying the training, plus the paid time of employees and trainers. Gains are the savings measured on your four numbers. Here’s an example:

Example: training on one production job, over one year
Building the modules and sheets− $6,000
Paid time for employees and mentor− $4,000
Total cost− $10,000
Scrap reduced by $500 per month+ $6,000
6 hires autonomous 2 weeks sooner+ $12,000
Total gains+ $18,000
ROI: (18,000 − 10,000) ÷ 10,00080%

Fictional example, to illustrate the calculation. Use your own data. The value of a week of autonomy gained depends on your production.

In this example, every dollar invested returns $1.80 in one year. And this calculation doesn’t even count accidents avoided.

Set up measurement in 5 steps

1

Pick one training

Start with a single course, on a job that causes problems.

2

Take a baseline

Record your numbers for the three months before the training.

3

Deliver the training

Keep a record: who was trained, when and on what.

4

Measure afterwards

Record the same numbers for the following three months.

5

Present the result

One page, with before, after, cost and gain. That’s your case for the next training.

Traps to avoid

✅ Do

  • Measure before you start
  • Compare the same job, over the same period
  • Count employees’ paid time in the costs
  • Stay conservative when estimating gains
  • Share results with the team

❌ Avoid

  • Measuring only the participation rate
  • Changing several things at once
  • Inflating gains to look good
  • Giving up after one month
  • Forgetting the cost of not training

The cost of not training

The real calculation compares two scenarios. Training has a cost. Not training has one too, often hidden:

  • scrap and rework that repeat every month;
  • an accident, with its stoppages, investigations and impact on the team;
  • a new hire who leaves after six weeks, and everything starts over;
  • an expert who retires with knowledge never passed on. See our 90-day plan for a key employee’s departure.
Training isn’t an expense when you can show what it brings in.

If your payroll reaches $2 million, your eligible training spending also counts under Quebec’s 1% law. Better to invest that amount in measured training than pay it into the Fund. And to track everyone’s autonomy, a skills matrix is the ideal tool.

Frequently asked questions

How long before seeing results?

It depends on the number. Errors and scrap often change within a few weeks. Early departures and autonomy take several months of tracking.

How do I put a figure on an accident avoided?

Use the real costs of your past incidents: lost hours, replacement, managers’ time, production stoppage. Stay conservative. If it’s too hard to estimate, leave it out and mention it as an additional gain.

Do I need software to measure?

No. A spreadsheet is enough to start. A training platform does, however, help track automatically who was trained and when.

What if the result is disappointing?

That’s useful information too. The training may be poorly targeted, too theoretical or not followed up on the floor. Adjust it, then measure again.

Want training whose effect you can measure? Witeach builds training targeted to your jobs and automatically tracks who is trained, so you can show the results.

Free succession plan →Talk to an expertExplore the platform

Share this article

What if we preserved your experts’ knowledge?

Every retirement takes years of experience with it. Let us talk about how to capture and pass on that knowledge before it leaves.

Let us discuss your needs