Most Improved: Learner

It measures learning improvement over a recent baseline, not raw volume alone. The template uses verified learning actions such as course.completed and quiz.passed, then ranks entrants by most improved performance. That makes it a good fit when you want to reward progress, not just the people who already score highest.

Use it for teams where training completion matters and some learners have already plateaued. It works well for onboarding cohorts, compliance refreshers, product training, and skill-building programs. It is especially useful when you want newer or slower-starting learners to have a fair path to win.

The template is set up for a 21-day duration, which fits a short learning sprint. That cadence is long enough for people to complete a few modules and show progress, but short enough to keep attention high. If your training cycle is longer, clone it into repeated rounds rather than leaving it open-ended.

A normal leaderboard rewards the highest raw totals, which usually favors people who started strongest. Most improved compares each entrant to their own recent baseline, so progress is what matters. That makes the contest more motivating for mixed-skill groups and less punishing for people who are starting behind.

Yes, but keep the prize framed as points or recognition rather than pressure. Because the scoring is based on verified system actions, it stays tied to real completion events instead of self-reported effort. If the training is mandatory, make sure the contest supports participation without creating a coercive environment.

The biggest mistake is using raw completion counts when the goal is improvement, because that can still favor the most active learners. Another pitfall is running it too long, which turns a contest into a leaderboard. Also avoid mixing in manual scoring for learning actions that the platform already tracks, since that weakens trust in the standings.

Yes. You can keep the most improved goal and adjust the prize type, point award, or contest messaging to match your program. If you want to emphasize participation over competition, you can also pair the standings with badges or recognition so the contest feels encouraging rather than cutthroat.

It is designed around verified learning events, so it works best when course completions and quiz passes are already captured in your system. That reduces manual entry and helps prevent Goodhart's-law problems where people game the metric. If your learning data lives elsewhere, map those events into the contest before launch.

Announce that the contest rewards personal improvement, not absolute rank. Explain that entrants are competing against their own baseline and that the format is meant to give everyone a fair shot. It also helps to keep the standings private or opt-in if your audience is sensitive about public ranking.