Slides about software don’t teach software. So this drops you into a mock register to process an actual return, with a coach explaining each step the first time through. Then you run it again alone. If the second run is clean, the training worked.
{{ screenPrompt }}
{{ coachText }}
The coach is quiet now. The policy card below is all the support you get, exactly like the sales floor.
In the shipped version this runs as a Storyline 360 try-mode simulation with screen-accurate captures of the client's POS; results post to the LMS per attempt, per mode.
I built two modes because the first run teaches and the second one measures. The coach narrates the reasoning behind each click, then goes quiet. Whatever survives the silence is what was actually learned.
I let wrong clicks happen instead of blocking them. Each plausible mistake gets its own consequence text, because nobody pays closer attention than in the second after getting something wrong.
I kept the policy card on screen in both modes because it exists at the real register too. Training people to work without the reference they will always have is training for a job that does not exist.
The coached run works inside the zone of proximal development, with expert reasoning narrated at each step. The support then fades on purpose. Competence is what remains when the scaffold comes down.
Errorful practice with corrective feedback outperforms errorless instruction. I made every wrong click a plausible one, so the mistake and its consequence text land at the moment of highest attention.
People perform best in the conditions they practiced in. Practice happens inside the register interface with the same policy card they will have on the floor, so nothing about the real task is novel on day one.