How to measure behaviour change in soft skills training

In L&D, we often measure what’s easy — completion rates, quiz scores, attendance. The problem? None of those tell you if behaviour actually changed.

If your training is meant to improve how people act in the real world, you need to measure differently.

Step 1 — Define the behaviours you want to see

Before you launch training, be clear on:

  • What will people do differently after this?

  • How will that look in their day-to-day work?

  • Which business outcomes will it impact?

Example: “Handle customer complaints with empathy, leading to fewer escalations.”

Step 2 — Capture baseline data

You can’t measure improvement without a starting point. Use:

  • Observations from managers

  • Existing performance data (e.g. complaint rates, safety incidents)

  • Simulation or roleplay results before training starts

Step 3 — Use realistic practice environments

If your training is a one-way presentation, you can’t measure behaviour.
Tools like AI-powered roleplay or simulations let you:

  • Track decision-making

  • Analyse priorities and response times

  • Identify common mistakes

Step 4 — Compare post-training performance

Measure the same way you did at baseline:

  • Repeat simulations

  • Gather fresh performance data

  • Ask managers and peers for feedback

Look for patterns: improved decision quality, faster resolution, better compliance.

Step 5 — Link it back to business results

When behaviour improves, you should see shifts in operational metrics. Examples from our projects:

  • 35% reduction in at-risk driving events

  • 99% onboarding engagement with measurable skill gains

  • $12m ROI through more efficient project delivery

Bottom line:
If you want to prove training works, stop measuring completions and start measuring behaviours. When you can show that link, it’s no longer “L&D spend” — it’s a business investment.

CTA: See how our process builds measurable behaviour change into every project

Totem Learning