Testing and improving
Review with an expert, pilot with a few learners, then use Analysis to fix what doesn't work.
No lesson is perfect the first time. The good news: with a short review before you send it and a look at the results afterwards, most problems are easy to find and fix.
Before you publish
Go through it as a learner
Open Preview and take the lesson on your phone, from start to finish, answering every question. You'll notice cards that are too long, images that are too small and questions that read differently than you meant.
Review with an expert
Ask someone who knows the topic well, ideally someone who does the job every day, to check the content and especially the questions. Use comments on the cards: they can point to a card, and you'll see what's still open.
Ask them specifically:
- Is everything correct, and is anything important missing?
- For each question: is the marked answer the only right one? Could you argue for another?
- Are the situations realistic? Would people really face them?
Pilot with a few learners
Send the lesson to 3–5 people from the target group before you send it to everyone. Then ask them:
- What was unclear or confusing?
- Which question did you hesitate over, and why?
- What will you do differently at work?
Five people find most of the problems. It costs a day and saves hundreds of frustrated learners.
After you publish
Use Analysis
Analysis (lesson menu ⋯ → Analysis) walks through the lesson card by card and shows:
- Questions to check: questions where fewer than 80 % of learners answered right the first time.
- How often each option was chosen, and the most common mistake.
- Cards where many learners stopped.
A question most people fail is usually the question's fault
When many learners get a question wrong, look at the question before you blame the learners:
- Most mistakes go to one wrong option? That option may actually be right too, or the right answer may be marked wrong, or the cards before suggest it. Read the question as if you'd never seen the lesson.
- Wrong answers spread across all options? People are guessing. The topic probably isn't explained well enough before the question, or the question asks about something the lesson doesn't cover.
- Success under 60 % almost always means something to fix. Between 60 and 79 %, it's worth a look.
A question everyone gets right isn't necessarily good either. It may be too easy to check anything. In that case, make the wrong options more tempting, or ask about applying the rule instead of recalling it.
Cards where people stop
If many learners leave the lesson at the same card, it may be too long, confusing, or a required question they can't answer. Shorten it, split it, or move it.
Fixing a live lesson
You can edit a published lesson at any time. Keep in mind that the first attempt counts and repeats don't change results, so people who already took the lesson keep their score. Fix questions early, ideally after the pilot. If you move or add cards, older attempts may not match the new order in Analysis.
Make it stick
A lesson isn't finished when it's published. What people remember a month later depends on what happens next:
- Review lesson: a few days after the main lesson, send a short Review & recall lesson: 6–10 questions mixing the topics.
- Spacing: plan a sequence such as main lesson → review after 2–3 days → scenario after a week.
- Ask about practice: a form like "Where have you used this since?" in the review lesson shows you whether the lesson changed anything at work.
- Talk to managers: what people practise in a lesson sticks when their manager asks about it the next week.