Gamification in corporate training

Gamification is cheap to add, easy to demonstrate to a steering committee, and almost always evaluated on the one number it was designed to move. Completion rates go up. Whether anybody behaves differently on Monday is a separate question that rarely gets asked in the same meeting.

We design and build custom learning games for clients, which mostly means we have watched gamification fail in specific, repeatable ways. This guide is the version we would want if we were buying: what each mechanic incentivises, when scoring makes training worse, and what to specify if the objective is judgment rather than attendance.

What each mechanic actually rewards

Every mechanic is an incentive, and people optimise for incentives with far more creativity than anyone plans for. Before choosing one, finish this sentence honestly: this mechanic will make people do more of ______. If the blank is not the behaviour you were hired to change, the mechanic is decoration.

Notice that only the last row rewards anything you would put in a competency framework. That is not an accident: the mechanics easiest to bolt onto existing content are the ones furthest from the behaviour you care about.

The leaderboard problem

Leaderboards are the first thing requested and the least examined. A ranked list does not create competition so much as publish who was already ahead, usually for reasons unrelated to the training: prior knowledge, free time, comfort with the format. The people at the top were going to engage anyway. The bottom third read the board once, conclude the race is settled, and disengage more thoroughly than if you had never scored them.

If you want the mechanic anyway, three modifications make it survivable: rank teams rather than individuals, reset often enough that a bad start is recoverable, and publish improvement rather than absolute score.

A leaderboard does not create competition. It reveals who was already winning, and quietly tells everyone else that this is not their event.

When gamification makes training worse

This is the section platform vendors do not write. Six situations where a scoring layer actively damages the programme, and most L&D teams will recognise at least two.

1. The work already matters to people. Attaching a small extrinsic reward to a task someone finds meaningful can crowd out the original motivation, an effect studied for decades as the overjustification effect. Attention paid to points comes out of the attention that was on the work.
2. The subject is serious and the mechanic trivialises it. A badge for completing a module on harassment, bereavement or a fatal safety incident does more harm than no badge at all. It tells people the organisation regards the topic as a box, and they will treat it accordingly.
3. You need honesty rather than performance. The moment a response is scored, people answer to be seen answering well. Anything that depends on psychological safety, disclosure or admitting a gap gets worse under scoring, not better.
4. The score becomes the target. Goodhart's law arrives on schedule: any measure that becomes a target stops being a good measure. Score completions and you will get completions, including the ones achieved by leaving a tab open over lunch.
5. The mechanic is trivially gameable. If the number can be moved without doing the underlying thing, somebody will find that route within a week and tell everyone else. Assume your most competitive participant is also your most inventive one.
6. The group is small or senior. Below a dozen people, a skilled facilitator with a whiteboard beats software. And a room of directors will opt out of anything that feels like school faster than any other audience you will meet.

None of this is an argument against games. It is an argument against decorating a course with scoring and calling the result a game.

Gamification is not the same as a learning game

• Gamification takes existing content and adds a reward layer on top: the content is unchanged, the incentive is new. The subject of the sentence is still a module.
• Game-based learning makes the game the content. There is a situation, a decision, a consequence, and the material is learned by acting inside it rather than reading about it first.
• Simulation goes one step further: the model behaves like the real system, so the consequence of a decision is generated rather than scripted. Expensive, and worth it when the real system is expensive to get wrong.

Most disappointing results come from buying the first and expecting the second. If the objective is knowledge, a policy or a product spec, gamification of the first sort is a cheap and reasonable tool. If the objective is judgment, it will not touch it, because judgment is not a completion problem.

What to do when the objective is judgment

When the behaviour you want is choosing well under pressure with incomplete information, the game has to be the training rather than a wrapper around it: a scenario with real ambiguity, decisions made by teams, a facilitator controlling the pace, and a debrief that turns what happened into what changes.

Two of ours run this way. HEARTWARE is five rounds over ninety minutes on judgment in the AI era, written for new joiners: teams decide what to trust, what to verify and what to own, and question the characters in free text rather than picking from a menu. Operation Keystone is a spy-thriller induction for two hundred new joiners in sixteen teams, deliberately hybrid, with ciphers inside the app and physical challenges run by facilitators off-screen. In both, reveals land on every phone at once so nobody races ahead of the conversation. The detail is on our gamified learning apps page, including the cases where we tell people not to commission one.

Something you can run on Monday: the one-decision round

You do not need software to test whether scored decisions beat scored completions. Forty-five minutes, eight to forty people in teams of four or five, one slide and a timer.

1. Write one real decision from your world. Three defensible options and no clean answer. A supplier missed a deadline; two teams claim the same budget; a customer wants what the contract does not cover.
2. Withhold something. Give each team a slightly different fragment of the background and do not tell them the fragments differ. Ten minutes to decide.
3. Score the reasoning, not the answer. Each team gets ninety seconds to state their choice and the one fact that would have changed it. You award points for the second half only.
4. Reveal the missing pieces. Show the full picture. Two minutes for teams to say whether they would change their decision, and why they never asked for what they did not have.
5. Debrief question: which of you knew you were missing information and decided not to go looking for it, and what does that habit cost us here? That is the whole exercise. The scoring exists to make people commit to a position they then have to defend.

Run it once with points on speed instead of reasoning and you will see the difference in the room within four minutes. That comparison beats any vendor deck.

Measuring it without inventing numbers

Corporate gamification is drowning in fabricated statistics, the most quoted being the learning pyramid: people remember ten per cent of what they read, ninety per cent of what they do. The percentages have no traceable source, the numbers are suspiciously round, and the institutions they are attributed to have said they cannot find the research. Experiential learning is defensible on better grounds than a fake statistic.

Measure it properly instead. The Kirkpatrick model gives you four honest levels, and gamification tends to produce beautiful results at level one and nothing at level three unless the design earned it. Pick one behaviour, agree how you would see it in the work, and check at ninety days.

A short specification you can hand a supplier

1. The behaviour we want changed is ______. Written before any mechanic is discussed. If it cannot be finished, stop here.
2. Individual or team scoring, and why. Team scoring surfaces how the group actually works; individual scoring surfaces who already knew.
3. What advances the session: a facilitator or a timer. Say which, out loud. It determines whether you are buying a session or a product.
4. Who runs the debrief, and for how long. If the answer is nobody, delete the game and write a good document instead.
5. What the game produces that a facilitator can use. Decisions, reasoning, timing, who went quiet. A final score is not evidence.
6. How it is reused. Per intake, per quarter, once. This is where the economics stop being frightening or turn out never to have worked.

If you want to feel the difference before commissioning anything, The Weekly Challenge is ten minutes of ours, free, ending in the same kind of debrief we build into paid work. Play it, then look at your current e-learning leaderboard and ask which one you would still remember in a month.

Tour De Force runs this as live, experiential training for teams worldwide — online and in person. Talk to us, or play Gamified learning appsThe Weekly Challenge to see the method in ten minutes.