The open-world shift

The problem was never a lack of tools

Walk into almost any enterprise revenue organisation in 2026 and you will find more sales technology than at any point in history: CRM, content library, conversation intelligence, coaching tools, AI copilots, forecasting models, and usually a few point solutions nobody remembers approving. What you will not reliably find is enough people who can actually run a complex, multi-stakeholder enterprise deal from first call to signature. That gap, a shortage of practised skill rather than a shortage of software, is the real bottleneck in B2B sales today, and it has been quietly misdiagnosed for most of the last decade.

The industry instinct, when quota attainment slips, has been to add more content: another battlecard, another onboarding module, another hour of asynchronous video. That treats the problem as an information deficit. It is not. It is a repetition deficit. Reading about how to handle a procurement objection and having handled one under pressure are not the same kind of knowledge. Game designers understood this long before gamification was a boardroom word: a player does not learn a difficult level by reading the manual twice. They learn by attempting it, failing, noticing exactly where and why, and trying again with stakes and feedback tight enough that the lesson sticks. That is a fairly precise description of how an enterprise account executive learns to run a deal too. The industry is only now building infrastructure that takes this seriously.

Correcting the timeline, because the dates matter

The popular story usually gets the dates wrong in one direction or the other: either treating gamification as a decades-old loyalty punch-card idea, or treating it as an app-store invention. Neither is accurate.

The term gamification was coined in 2002 by games researcher and consultant Nick Pelling, but it sat almost unused for the better part of a decade. It did not enter mainstream business vocabulary until 2010, the year the verb gamify was coined and the first Gamification Summit was held in San Francisco. Sales enablement, as a named discipline distinct from general sales training, followed a strikingly similar arc: widespread industry adoption around 2008, and the first wave of dedicated enablement platforms scaling from roughly 2010 onward.

That overlap is not a coincidence worth glossing over. Gamification and sales enablement, in the form both fields take today, are twin disciplines born in essentially the same stretch of years: not because either idea is that old in the abstract, but because that is when both stopped being curiosities and started being applied, deliberately and commercially, to enterprise revenue. Measured from that starting point, the arc really is approaching two decades by 2027. That is the honest version of a nearly twenty-one years framing: not the 2002 coining alone, which predates serious enterprise sales application by close to a decade.

Four eras, and what became table stakes

Looking back across that arc, the field breaks fairly cleanly into four eras. The useful way to read each one is not what was new, but what started as an edge and ended as a baseline expectation. That pattern is the engine of the story, and it is about to repeat again.

The psychology underneath, and why badges plateaued

The shift across those eras is not only technological. It tracks a paradigm change in the psychology being applied.

Early gamification mechanics (points, badges, streaks, leaderboards) are almost without exception applications of operant conditioning: external reinforcement to encourage a behaviour. It works, and the engagement literature is real. Extrinsic reinforcement of this kind also has a well-documented ceiling. Motivation research going back to self-determination theory shows purely extrinsic rewards are subject to novelty decay (the tenth badge motivates far less than the first) and can, in some conditions, crowd out intrinsic motivation to master a skill. Any sales leader who lived through the platform era recognises the pattern: leaderboards produce a burst in month one and are ignored by month six.

What is emerging now draws on a different body of psychology: flow (challenge matched to skill), deliberate practice (structured, feedback-rich repetition), and situated or apprenticeship learning (skill through realistic participation, not abstract instruction). None of this is about extrinsic reward. It is about designing an environment where a person can fail safely, get immediate and specific feedback, and try again: the loop game designers already used to teach difficult levels, decades before anyone applied the word gamification to a sales team.

That is the paradigm shift worth naming. Gamification first wave borrowed games aesthetics: points, badges, competition. The current wave is finally borrowing games underlying mechanic: safe, repeated, high-fidelity practice. That is a meaningfully different, and more defensible, claim. It is also why we argue elsewhere that gamification needs a system to design based on. It cannot live alone as sugar.

Why now: four converging, measurable forces

It is one thing to argue that simulation-based practice is psychologically sound. It is another to explain why enterprise sales is building this infrastructure now, in this three-to-five-year window.

The cost of building an adaptive counterpart has collapsed. A genuinely open-world practice environment needs a counterpart that reacts to what a rep actually does rather than a pre-written branching script. Andreessen Horowitz analysis shows large language model inference cost dropped by roughly a factor of 1,000 between late 2021 and late 2024. What was economically absurd (a persistent, memory-holding AI counterpart for every practice session) is now cheap enough to be standard infrastructure. This threshold, not a change in philosophy alone, is a primary reason open-world sales simulation is a 2022-onward phenomenon rather than a 2015 one.

The real cost of an unrehearsed mistake has gone up. B2B buying committees grew more complex over almost exactly the same window sales enablement existed as a discipline. Industry reporting tied to CEB and Gartner shows average enterprise buying groups growing from about 5.4 stakeholders in 2015 toward ranges of 8 to 13 in the mid-2020s. Every additional stakeholder is another way for an unrehearsed rep to say the wrong thing in a live, revenue-bearing conversation. When the committee was five people, on-the-job learning was more survivable. At eight to thirteen, the first real mistake on an actual deal is a risk-management problem, not a soft training concern.

The informal channel that used to teach this skill has narrowed. Before remote-first selling became standard, a meaningful share of how junior reps learned to run a deal was never written down. It was absorbed by sitting near, and occasionally sitting in on calls with, more experienced colleagues. That channel never needed a budget line. Its erosion under hybrid and remote work is widely observed in onboarding and ramp commentary, and it is the missing piece that explains why a formal practice environment became necessary rather than merely nice to have.

Software production itself is being commoditised, which raises the value of whoever can still sell it. AI-assisted development collapses the cost of building software. In consumer and indie software, distribution can often be bought or automated. Enterprise B2B does not have that option. An eight-to-thirteen-stakeholder buying committee is not won with a content engine. It is won by a skilled person navigating competing, sceptical stakeholders. As AI makes it cheaper to build a competing product, it does not make it cheaper to sell a complex one, so the relative value of a rep who can actually run that deal keeps rising.

Individually, none of these four forces would be enough to treat rehearsal infrastructure as a strategic priority. Together, they change the category the investment belongs to: from learning and development toward something closer to revenue risk management.

From levels to open worlds

This is where the gaming analogy earns its place. The first wave of gamified sales training, including much of what shipped in the platform era, was underneath the surface a fixed branching tree: a scripted scenario with a small number of pre-written paths, dressed up with points and badges. Structurally, that is closer to a fixed obstacle course than it might like to admit: the same order every time, mastery mostly meaning memorising the correct sequence.

What is different about the current generation is structural, not cosmetic. An open-world practice environment gives the counterpart (the simulated buyer) its own memory, goals, and reactions, so the same starting scenario plays out differently depending on what the rep actually does, the way an experienced buyer would. That is a meaningfully different design problem than writing a branching script, and it mirrors a shift games themselves went through: from level-based design toward worlds that respond to the player.

One concrete illustration of this category is what we build at Game Is Serious in Berlin: Deal IQ for seat-specific deal simulation, Team IQ for inside-pod soft craft, Mirror and Transfer as diagnostic and advanced coaching, all feeding a manager-facing Revenue Command Center. Two things about that approach matter independent of the feature list. First, the differentiator is the counterpart realism and the specificity of the evidence, not the presence of points or badges. Second, we stay claim-safe: no invented ROI percentages on this page, evidence-bound language in a category where inflated engagement statistics are common.

A less obvious implication extends past training into hiring. The standard way to evaluate whether a sales candidate can run a deal is still an unstructured interview about past performance. A structured, frozen simulation scenario, the same one every candidate faces, generating comparable behavioural evidence, is a categorically better instrument for that judgment: a work sample of the actual skill, rather than a proxy for it. That is a structural consequence of treating simulation as evidence infrastructure rather than engagement decoration.

What comes next

Extrapolating the pattern that has repeated across all four eras, a reasonably confident mid-term prediction follows. Within two to three years, AI-generated call coaching and basic simulated role-play will be table stakes, the same way leaderboards were table stakes by 2016 and content repositories by 2010. The edge will not disappear. It will keep moving toward fidelity: how convincingly the counterpart reacts, how safely and specifically a rep can fail and recover, and how directly practice evidence translates into a coaching conversation a manager actually has.

The broader lesson sits above any specific product category. Sales enablement has spent roughly twenty years re-learning, in a new vocabulary every few years, something game design worked out generations earlier: skill is not transmitted by reading or watching. It is built through repetition, with real stakes and honest feedback, until the response becomes intuitive rather than recalled. The industry next five years look less like more content and more like better-built worlds to practice failing in: quietly, safely, and often enough that by the time the stakes are real, the rep already knows what to do.

If you want the system-before-sugar argument, read Design practice people will enter. If you want the taxonomy of AI sales training bets, read The five bets. If you want the design-craft argument about session tempo, read The pacing problem. If you want the short search landings, use the gamification and AI buyer simulation guides. Here the line is simpler. The open-world shift.

Frequently asked questions

References

• The History of Gamification (Gamify.com)Article
• The History of Sales Enablement (Aspiration Marketing)Article
• LLMflation: LLM Inference Cost Is Going Down Fast (a16z)Article
• B2B Buying Committees: Why 8-13 Stakeholders Changed Enterprise SalesArticle