Perspective · Berlin

One world, three layers: what a sales team needs before it enters a practice world twice

· 4 September 2026 · 7 min read

One world, three layers.

Key claims

  • A game people enter twice is never only the rules. It is the rules, the place, and what you did there last time.
  • Products got cheap. Selling did not. The bottleneck of the company moved from the engineering floor to the sales floor.
  • Slides, videos and multiple choice were cool once, the way calling a taxi dispatcher was cool once.
  • Roleplay without a world is a scene, and a scene is watched once.
  • If you can swap the logo and the room still works, you are not in the customer's world yet.
  • An open world is not more branches. It is a counterpart that remembers what you did.
  • Evidence a manager can defend is part of the design, not a report bolted on afterwards.

One world, three layers.

A street, a phone, and a route nobody repeated

Years before Berlin, in Tehran, a team of colleagues stood on a street corner with one phone between them and a riddle that would not open. A second team from the same company was three blocks away on a route nobody had planned for them. The city hunts of those years, and the board games we published as Farbood, taught me something I did not have words for until much later.

The city was already there. That was the first layer: a real place with real rules, nothing to invent. The second layer was ours to build: the case, the clues, the stakes, tuned to whoever was playing that day. The third layer belonged to the players. Two teams started at the same door and came back with different stories, because the route was theirs. Eight years, 200 sessions and 60 organisations later the pattern held every time. A game people enter twice is never only the rules. It is the rules, the place, and what you did there last time.

Why sales, why now

A decade ago the engineers were the heart of a company and the sales floor waited for what they shipped. That order is flipping. Products got cheap to build. AI and a weekend of vibe coding now produce what used to take a funded team a year. Selling did not get cheap. The bottleneck moved from the engineering floor to the sales floor, and far more people have to learn to sell well, in less time, with a calendar that is already full.

Most enablement still answers that with the formats that were cool in 2010: a slide deck, a video, a multiple-choice quiz, a completion certificate. They were cool once, the way calling a taxi dispatcher was cool once. Nobody calls the dispatcher now. They open Bolt. Adults protect their attention from anything that feels like homework, and the question worth asking is not how to push learning harder. It is what people would pull.

The pilot never stops flying the simulator

Learning by doing is the old answer and it still works, because a mistake you made yourself hurts in a way a slide never will. The pilot learns in the simulator and is still flying it in the last year on the job, not only the first. The surgeon rehearses. Sales, the profession where one bad conversation costs a quarter, has mostly rehearsed by reading.

The market has noticed, and the new fashion is the short scene: one AI buyer, one call, one score. It is interesting the first time. By the third time it is theatre, and the retention numbers say the same. A proper conversation is hard. Ask my wife how many years it took me. It is harder for a language model, however well it is prompted, when it has no memory of who you are and what you did last week. Roleplay without a world is a scene, and a scene is watched once.

Layer one: the job, whole

The first layer of Deal IQ is the job as it actually runs, not a call dressed up as training. Territory planning, account research, pipeline, the live conversation, the debrief. By the time a player faces the buyer, the deal has a history the player made, and that is the only reason the conversation carries weight.

The framework belongs to the seat. SDR and BDR rooms run SPICED. Account executives are scored on MEDDPICC. Customer success practises expand and renew. One acronym forced across every seat is how enablement loses its veterans. The base world is the shared craft floor: same spine, judged the way that seat is actually judged.

Layer two: the company world

In the tabletop years this layer was literal. The rules travelled from client to client; the cards never did. They were printed with that company's market, its products, the objections its people actually heard. Sixty organisations across a dozen industries, and not one asked for the generic deck twice.

In Deal IQ the second layer is built the same way, mostly from public data first: the industry, the products, real accounts and the signals they leave in the open. Then it deepens with what only the customer can give: ICP, guidelines, strategy, the motion the team really runs. The test is simple. If you can swap the logo and the room still works, you are not in the customer's world yet.

Layer three: the personal open world

At a Dungeons and Dragons table the game master says three weeks pass and nobody blinks. Then a single door takes twenty minutes, because behind it is the moment that matters. That is the mechanism I borrowed. Alex, the game master in Deal IQ, folds the empty weeks of a deal cycle into a sentence and slows the beats that deserve full pressure, so a cycle that takes months fits a session someone actually finishes.

What makes the world personal is memory. The AI agent holds the whole context of what this player did until now, so when they ask for help the guidance is about their deal, not a tip from a library. Two reps start in the same account and leave with different scars, because the buyer remembers what each of them said. Tips stay assistive; nothing applies itself. The player keeps the pen. An open world is not more branches. It is a counterpart that remembers what you did.

What the manager sees

In 2020 the rooms went remote and a client asked the question that changed the company: can you measure what just happened. Ten years of enterprise rooms teach the same lesson. Managers do not want a feeling from a workshop. They want credible, data-driven evidence in the place they already manage from.

So the fourth thing is not a layer but a window. In the Revenue Command Center a manager reviews the evidence a run produced, then assigns the next mission or revises it. Review, then assign. Evidence a manager can defend is part of the design, not a report bolted on afterwards. Without it, the practice world becomes private theatre the moment the facilitator leaves.

Enter it twice

A sales team deserves the seriousness a pilot gets from a simulator: a world with the whole job in it, grounded in the company it sells for, paced for the person inside it, and visible to the manager who has to defend the quarter. The first layer makes it real. The second makes it yours. The third is why anyone comes back. One world, three layers.

Enter one deal practice. No login.

Frequently asked questions

What are the three layers of a practice world?

Layer one is the base world: the whole seat job from territory planning through the live conversation to debrief, judged with the framework that seat actually uses. Layer two is the company world: the industry, the products, and real public accounts, so the room could not belong to another vendor. Layer three is the personal open world: the path bends with what the player does, paced by a game master, with an AI counterpart that remembers.

Is this different from AI roleplay?

Yes. Roleplay is one conversation without a world around it, and it is interesting once. A practice world is a finishable job sequence grounded in the company, scored the way the seat is scored, visible to the manager, and personal because the counterpart remembers earlier moves. AI holds memory inside that world. It is not the whole design.

What does the game master do in a sales simulation?

The same thing a game master does at a Dungeons and Dragons table. Alex says three weeks pass and nobody blinks, then slows a single conversation to full pressure because that is the moment that matters. A deal cycle that takes months fits a session someone actually finishes.

Who designs Game Is Serious practice worlds?

Sohrab Mostaghim, founder and lead designer of Game Is Serious in Berlin, with the studio team. He published board games as Farbood and ran city hunts and bespoke company games for a decade before designing practice worlds for revenue teams. The rooms are built as serious games, not as LMS modules with badges attached.

Is startup gamification badges or a practice system?

Badges alone are sugar. Startup gamification that lasts is a practice system people enter twice: a world with the whole job in it, grounded in the company, with a personal path. Points are optional. The world is not.

References