Serious games at work: why people learn by playing
Serious games at work: why games build data and AI habits, what makes a good serious game, and how to measure what your teams actually learn.
A serious game is a game designed to teach, not only to entertain. At work, its main job is to build habits: decisions people have to make quickly and correctly without reopening a course every time. Data and AI literacy is a natural fit, because it comes down to repeated decisions: hand this task to AI or not, trust this answer or not, keep this row of data or fix it.
Why a game rather than a standard e-learning module?
A module that presents rules and ends with a quiz passes on information. A game makes people decide. The difference comes from a few simple mechanisms:
- Retrieval practice. Recalling a rule in order to apply it anchors it better than reading it again. A game has players apply the same rule dozens of times, in changing situations.
- Immediate feedback. Right after the decision, the player sees the consequence and the reason. The mistake becomes an example at the moment it is most useful.
- Gradual difficulty. Early rounds establish the rule, later ones add edge cases and traps. Players progress without being discouraged.
- Safe failure. In a game, being wrong costs nothing, so people dare to test their understanding, which they rarely do in front of a real customer.
- A reason to replay. A score, a streak, a leaderboard among colleagues: people start another game on their own, which multiplies the repetitions.
What makes a good serious game?
Not every game teaches. A few criteria help you judge one:
| Criterion | What to look for |
|---|---|
| Short | a game fits in a break: 5 to 10 minutes, not an hour |
| Realistic | cases look like the players' actual work, not a general-knowledge quiz |
| Explicit rule | after each decision, the reason is stated in a line or two |
| Skill-level measurement | the result shows what is mastered and what is not, not just an overall score |
| Never punitive | a mistake explains; it does not humiliate or block progress |
| Accessible | playable with a keyboard, readable by screen readers, not dependent on sound |
Maketools games follow these criteria: solo games of 4 to 10 minutes, an explanation after every answer, named skills, full keyboard play, text readable by screen readers.
How do you measure what teams learn?
A serious game only earns its place at work if you can track its effect. There are three levels of measurement:
- Participation: who played, how many plays, what completion rate. This is the first evidence that a programme has been rolled out.
- Mastery by skill: for each skill, the share of correct answers. This is what steers the next step: if "detecting bias" stays low, that is where to focus.
- Progress: how scores change over time and by group.
In Maketools, the score is recalculated by the server from the answers, so it cannot be inflated from the browser. Reports show participation, completion, scores and skills, and can be exported to CSV or PDF, or sent to your LMS (see LMS exports).
Play without an account, then customise
Two complementary uses:
- Discover: catalogue demos can be played without an account. They are useful for getting a feel for the games, or for an open awareness session (see events).
- Deploy: an organisation forks a game, creating its own copy of the configuration, adapts the cases, rules and colours in the editor without code, then publishes it for its members (see the studio).
Adapting cases to the organisation's context often makes the difference: a game about personal data that mentions your own tools and teams is more convincing than a generic example.
Where the format stops
A game does not do everything. It trains habits, but it is not enough to explain a complex internal policy or to teach a specific tool. It works best inside a wider programme: a kick-off, a short policy, conversations between colleagues, then regular plays. The 30-day AI literacy plan shows how to combine them.
Try it
The simplest way is to play: open the catalogue and start, for example, Hallucination Hunter or AI or Not AI?, no account needed.