Post-training survey questions that measure more than happiness

Four levels, four different surveys

The Kirkpatrick model splits training evaluation into reaction, learning, behaviour and results. Most organisations only ever measure the first one, then wonder why training budgets get questioned. The fix is not a longer survey — it is asking different questions at different times.

Level 1 — Reaction (ask immediately)

• How useful was this session to your day-to-day work? (1–5)
• How clearly was the material delivered? (1–5)
• Was the pace too slow, about right, or too fast?
• Was the balance of theory and practice right?
• How likely are you to recommend this session to a colleague? (0–10)
• What was the single most valuable part?
• What should we have spent more time on?

Level 2 — Learning (ask immediately)

• How confident are you applying this now, compared with before? (1–5, asked as a pre/post pair)
• Name the three steps of the process we covered.
• Which concept is still unclear?
• Describe a situation from your own work where you'd use this.
• What would you need in order to feel fully confident using it?

A short knowledge check beats a self-rating. Self-reported confidence tends to peak straight after training and fall once reality intervenes.

Level 3 — Behaviour (ask 30 days later)

• Have you used anything from the session in the last four weeks? If so, what?
• How many times, roughly?
• What made it easy to apply?
• What got in the way — time, tools, manager support, or something else?
• Has your manager discussed the training with you since?
• What would help you use it more?

Level 4 — Results (ask 60–90 days later, with the business)

• Which metric was this training meant to move, and has it moved?
• Can the team point to a specific outcome they attribute to the session?
• What is the cost of the problem the training was meant to solve, now versus before?
• Would you commission this session again for another team?

Rules that keep response rates high

• Under 90 seconds for the immediate survey. Six to eight questions maximum.
• One open text box per survey — more than that and quality collapses.
• Ask in the room, not by email, for level 1 and 2.
• For the 30-day follow-up, keep it to four questions and say up front that it takes one minute.
• Always tell people what changed because of the last cohort's answers. Closing the loop is the cheapest way to raise your next response rate.

Turning answers into action

Pick one change per session and name it. Feedback that produces a list of twelve improvements produces zero. PulseCheck's AI read-out groups responses into themes and proposes concrete actions, so the shortlist writes itself and you can compare it against your previous sessions.