You finish a course and score 87%. Pretty good, right?
Maybe.
What if everyone else scored 96%?
Suddenly, that exact same 87% feels very different.
That’s the problem with a lot of training feedback. We give people a number, but we don’t always give them enough context to know what that number means.
What Happens When You Add a Benchmark?
A recent field experiment published in Management Science looked at employees taking a voluntary online training program on agile project management and OKRs.
One group saw only their own test scores. The other saw their scores plus how their performance compared with other employees, displayed anonymously in deciles.
The difference got my attention.
Among employees eligible to continue, the odds of starting the next module were 4.35 times higher for those who received relative performance information.
Even more striking, just 17.5% of the score-only group finished all four modules, compared with 66.7% of the comparison group.
Same training. Same tests. The big difference was that one group had a meaningful reference point.
A Number Needs Something to Push Against
Think about the difference between these two messages:
You scored 87%.
You scored 87%. That puts you in the top 20% of participants.
Same number. Very different signal.
The study builds on social comparison theory: when objective standards are fuzzy, people often look to others to help judge their own ability.
An 87% score tells you what happened. Knowing where 87% sits relative to a useful benchmark helps tell you what it means.
That distinction matters in L&D because we throw isolated numbers at people all the time. Quiz scores. Assessment results. Completion times. Practice scores.
Then we assume the learner knows how to interpret them.
Often, they don’t.
Please Don’t Turn This Into Another Leaderboard
The takeaway here is not, “Excellent! Let’s rank everyone.”
In this experiment, the comparison was private. Employees knew where they stood, but they couldn’t see anyone else’s individual score.
The researchers also used deciles rather than a winner-take-all ranking, giving participants a clearer sense of their position and potentially a realistic next level to aim for.
That suggests a much more useful design question:
Could we give learners a meaningful benchmark without turning the experience into a competition?
You might tell someone that most first attempts land between 70–80%, that their result is higher than 75% of participants, or that people in similar roles usually complete a task within a certain range.
You’re not creating winners and losers. You’re simply making the feedback easier to interpret.
There’s an interesting wrinkle here, too. The effect was stronger among people who were more competitive, which probably won’t shock anyone.
But the researchers didn’t find evidence that the effect depended on whether people had previously performed particularly well or poorly. In this setting, comparative feedback didn’t appear to chase off the people at the bottom or cause the people at the top to stop trying.
Try This
Pick one piece of feedback from something you’re designing right now—a quiz score, simulation result, practice activity, whatever—and ask:
Does this number actually mean anything to the learner?
If the answer is “not really,” experiment with adding a useful reference point.
Not everywhere. Not publicly. And not as some giant gamification initiative.
Just enough context to help someone answer the question they actually care about:
How am I doing?
One important caveat: this was one field experiment with only 79 employees, at one company, in one voluntary training program. The researchers also couldn’t observe exactly how participants interacted with the learning content itself.
So this isn’t a universal rule.
But it is a promising finding worth testing in your own context.
The Bottom Line
Feedback without context is often just a number.
If you want learners to know how they’re doing—and perhaps give them another reason to keep going—don’t only tell them their score. Help them understand what that score means.
Sometimes the benchmark is the part that makes the feedback useful.
Go Deeper
The Impact of Relative Performance Information on Employees’ Persistence in Voluntary Training
The research behind this week’s issue. A real-world experiment tests what happened when employees could see not just their scores, but where those scores stood relative to others.
Less Information, More Comparison, and Better Performance
A fascinating companion study. In this field experiment, relative performance information alone produced stronger performance effects than showing people both absolute and relative measures—an intriguing reminder that more feedback isn’t automatically better feedback.
See why less information sometimes works better →
Relative Performance Feedback: Effective or Dismaying?
Here’s the useful counterpoint. Comparative feedback didn’t help everyone equally, and some lower-ability participants performed worse depending on how the comparison was designed.
The benchmark matters.
See when comparison can backfire →