How to create a B2B serious game in 9 steps

What distinguishes a real B2B serious game from a disguised marketing game

There is massive confusion over the term serious game. A significant part of the devices labeled "serious game" in business are in fact marketing materials with a playful appearance - not educational tools. This distinction is not semantic, it conditions the result. Marketing support optimizes the memorability of a brand. A serious game optimizes a change in behavior, the acquisition of a skill or the consolidation of knowledge.

Three operational criteria make it possible to decide. First criterion: the learning objective is measurable. “Making awareness of our new CSR policy” is not a measurable objective – it is a communication intention. “Allowing 80% of employees to identify the 5 energy-saving actions at their workstation” is measurable. Second criterion: the game produces an explicit feedback loop. The player sees, in each turn, the impact of their choices - this is what transforms the experience into learning. Third criterion: post-game evaluation is integrated into the system. A quiz, a structured debrief, a questionnaire on D+30: without this component, we have no trace of effectiveness.

Of the 33 B2B projects documented in our achievements (Enedis, SNCF Voyageurs, Veolia, France Travail, Red Cross, etc.), the three criteria are systematically present. On projects that are rejected in the framework, the missing criterion is almost always the measurable objective - proof that the rigor of definition is at stake from the brief.

Step 1 – Frame the learning objective and delineate it

The scoping stage generally takes 2 to 4 meetings depending on the maturity of the sponsor. The classic trap is over-scope: wanting to deal with "CSR", "cybersecurity" or "employee engagement" in a single game. These subjects are too broad for a single system - you have to break it up. About the project Watt for Enedis, the initial objective "to raise awareness of the energy transition" was reformulated as "to allow employees to identify the 6 concrete energy saving levers applicable in their daily lives". This reformulation saved the project: it made it possible to design a game system where each mechanic corresponded to an identified lever.

Methodologically, the grid to apply is as follows: objective → public → context of use → duration constraint → success indicator. As long as this grid is not measurably filled, we do not move on to the next step. This is also the stage where we integrate the industrial constraints - quantity, deadline, budget - which will structure everything else. See also our standard specifications for a B2B game.

Step 2 - Identify the audience and its usage constraints

A serious game is designed for a specific audience, not for “employees in general”. Three variables to document: level of prior knowledge (beginner, intermediate, expert), physical context of use (face-to-face training room, in office, on-site team, remotely), time constraint (15 minutes in a meeting, 1 hour in a workshop, half-day in a seminar). These three variables radically condition the final format.

The Right Reflex for SNCF Voyageurs The Right Reflex for SNCF Voyageurs, the audience was made up of field agents with a maximum of 20 to 30 minutes between two rotations. This time constraint prohibited the “90-minute complex board game” format and imposed a flashcard format playable in 15 minutes. About the project Emulation for France Travail, the public advisor in the agency had 1 hour in a collective workshop: the collaborative platform format was essential.

The trap: relying on the “ideal” audience described in the brief rather than the actual audience observed. A field visit or a focus group of 3-4 representative users upstream costs a day but avoids weeks of iteration.

Step 3 - Choose between physical, digital or hybrid format

The choice of format is not a question of fashion but of measured educational effectiveness. The data converges: the physical format has a significantly higher usage rate in digital format on B2B devices of less than 60 minutes, and a 30-day retention of 40 to 60 points greater. Conversely, digital outperforms physical on long devices (> 4 hours) or requiring individual monitoring.

Our recommendation by use case: for awareness raising, onboarding, team building, corporate culture (duration 20 min to 2 hours, face-to-face). for awareness raising, onboarding, team building, corporate culture (duration 20 min to 2 hours, face-to-face). Hybrid for medium professional training (4-8 hours) with digital theoretical component + physical practical workshop. Pure digital for certifications, long individual journeys, 100% teleworking contexts.

See our detailed comparison of physical vs. digital serious games with the benefits and drawbacks of each option by project type.

Step 4 – Design the game system (mechanics, narrative, aesthetics)

This is the most creative step - and the trickiest. Three bricks to be designed jointly: the game mechanics (rules, victory conditions, internal economy), the narration (universe, characters, emotional thread), theaesthetic (artistic direction, color code, typography). A common mistake is to design the aesthetic before the other two - we end up dressing an empty shell.

The methodological order: mechanical first (it carries learning), narration then (it carries commitment), aesthetic finally (it carries identity and perceived quality). On school publishing projects (Climate Fresco - Bordas, etc.), this order is strict. On creative agency projects, the order is often reversed - hence the long and costly iteration loops.

Useful tools at this stage: the game loop formalized on 1 A4 page, the mood board for aesthetics, the 200-word narrative synopsis. If these 3 deliverables do not hold up in cold reading, the concept is probably too vague to move into prototype.

Step 5 - Prototype in 'rough' mode (not pretty yet)

The rough prototype is the secret weapon of a mastered project. Printed on ordinary paper, cut with scissors, without complete artistic direction: the goal is to test the mechanics, not the aesthetics. Cost: a moderate envelope versus a moderate envelope for a semi-finalized prototype. Deadline: 2 to 4 days versus 2 to 3 weeks. On the supported projects, the rough prototype made it possible to identify 60 to 80% of mechanical problems on average - therefore avoiding as many costly loops in production.

Three pitfalls to avoid when rough prototyping. Trap 1: want to make it pretty - we lose the objective (test the mechanics quickly). Trap 2: test alone - you need at least 3 naive players. Trap 3: ignore observed friction under the pretext that we know the project by heart - the frictions seen by the naive player are almost systematically the frictions which will remain in production.

Step 6 - Test on a representative panel (user test)

The most neglected and probably the most profitable step in project ROI. A successful user test involves 6 to 12 representative players, over 2 or 3 sessions, with structured observation by a trained facilitator. Total cost: a moderate budget. What this avoids: a production of 5,000 to 20,000 units deployed in the field before discovering that a rule is misunderstood or that the duration is underestimated.

Criteria for a useful test: players representative of the final public (not the project team nor those close to them), simulated usage context (in a typical meeting room, not at home in the evening), structured observation (question grid, timer, verbatim notes). Indicators to follow: actual duration vs. announced duration, number of times the rule is reread, verbalized level of pleasure, ability to explain the learning objective at the end.

On the case Implike - Keolis, 3 test iterations reduced the duration from 75 to 45 minutes while increasing the memorization score to 30 days. Without this loop, the initial deployment would have had a usage rate 2x lower.

Step 7 - Iterate 2 or 3 times based on field feedback

Iterate does not mean “redo”. Iterate means: isolate the 3 or 4 priority friction points identified in the test, correct these points in the prototype, repeat a half-test (3-4 players) to validate. A well-conducted iteration takes 1 to 2 weeks. Three iterations generally cover 80 to 90% of the friction observable before production.

Methodologically, the golden rule: correct one variable at a time between 2 iterations. Otherwise we no longer know where the improvement or degradation comes from. To follow properly, a mini Excel table (iteration / friction / correction hypothesis / post-test result) is enough. This discipline distinguishes a professional approach from amateur development - and it shows in the quality of the final product.

Step 8 - Industrialize: pre-press, manufacturing, quality control

Industrialization is the stage where an integrated European B2B manufacturer like Craft Your Games takes control. Three sub-steps: pre-press (vectorization of fonts, colorimetric profile CMYK (printing color standard), overhangs, BAT (Good to Print, validation before printing) digital then paper BAT), manufacturing (offset or digital printing depending on the print run, shaping, finishing), quality control (tests EN71-1, EN71-2, EN71-3 by independent laboratory for games intended for minors or educational contexts).

The standard deadline for a B2B France project: 4 to 6 weeks after validation of the final proof, for a print run of 500 to 5,000 units. Compared to Asian sourcing (10 to 14 weeks + 3 to 5 weeks of transport + risk of delays of several weeks), this is 2 to 3 months saved. On tight schedule projects (CSR launch, dated HR event), this is what makes it possible to keep the commitment.

Step 9 – Deploy, measure, capitalize

Deployment is the step we neglect the most. Three levers to activate: internal animation (announcement email, facilitator kit for managers, 2-minute presentation video), usage monitoring (who received, who played, how many times, in what context) and a measure of impact at 30 days and 90 days.

On impact indicators, useful benchmarks: usage rate at 6 months (objective > 60%), spontaneous social sharing rate (objective > 30%), memorization at 30 days (objective > 70% on target questions), internal Net Promoter Score (objective > 30). These indicators feed into the ROI displayed to the sponsor - and the decision to redeploy or expand. On the cases Veolia And France Travail, these metrics triggered an add-on order 8-12 months after the initial launch.

Sources: AFNOR (games and toys standards) · Ministry of Labor (professional training page).

Reference guide: For the complete overview, consult our guide complete guide to serious games in business.