Cricket Nets for Sales Teams: Why Deliberate Practice Beats One-Off Workshops
Most enterprise sales training follows the same shape: a workshop, a role-play exercise with a colleague standing in for a prospect, a feedback form, and then nothing until the next annual refresh. In August 2026, Gartner published its first Market Overview for B2B Sales AI Role-Play Applications — confirmation that AI-simulated practice is now its own category, distinct from lecture-based training and from live coaching. That's a market signal, not a fad. The reason L&D budgets are moving here is simple: a workshop can tell a rep what to say once. It cannot give them anywhere to say it again before the deal that actually matters.
Why the one-off workshop doesn't hold
Hermann Ebbinghaus documented the forgetting curve in 1885, and a 2015 replication by Murre and Dros found the same shape holds up: sharp early decline in what's retained, then a slower fade. What determines how much of that decline applies to a specific skill isn't the quality of the original session — it's almost entirely whether the learner used the skill again afterward. A well-designed, well-received workshop can still fail to change what a rep does on a call three weeks later, for the simple reason that nothing in those three weeks asked them to retrieve and use what they learned. The workshop was never the problem. The absence of anything between the workshop and the next real deal is.
The cricket-nets model
Teachrity's founding framework for this is a cricket net. A batter does not learn to face fast bowling by sitting through a seminar on technique, and does not learn it by walking straight into a match either. They learn it in the nets: delivery after delivery from a bowling machine or a net bowler, in conditions close enough to a real innings to matter, with a coach adjusting stance and shot selection between balls. The nets are not the match — nobody is keeping score, and a mistimed shot costs nothing. They are also not a lecture. They are built specifically to be run again the same afternoon if the first attempt didn't land.
Map that onto a sales organisation and the pieces are already there, just not connected. The workshop is where a manager explains what good discovery or negotiation looks like. The live customer call is the match: performance counts, and a mistake has a real cost. What's usually missing is the net — a way to run the same difficult conversation, a price objection, a stalled procurement committee, as many times as it takes, with a coach's read on each attempt, before it happens for real.
What actually counts as deliberate practice
Not all repetition qualifies. K. Anders Ericsson's research on expert performance set out specific conditions for practice that actually builds skill: a well-defined task pitched just past the learner's current ability, immediate and specific feedback, and enough repetition to work through the same problem more than once. Two colleagues role-playing after a workshop rarely meets any of these conditions in practice:
A specific, evaluable scenario — not "practice your pitch," but a named buyer persona with a named objection
Feedback on that exact attempt, not a general debrief days later
Repeatable on demand, not scheduled around a manager's calendar
Difficulty calibrated to sit just past current skill, not a fixed script run the same way every time
Where AI simulation fits — and where it doesn't
This is the specific gap AI voice simulation is built to close. It isn't a replacement for the workshop or the manager — it's the net that sits between them. A rep runs a negotiation against an AI counterpart calibrated to a specific persona: a procurement lead who won't move on price, a champion who's gone quiet, a CFO asking for ROI proof. They can run that scenario as many times as it takes, get an evaluation on that exact attempt, and go again the same day, without booking a colleague's time. The workshop still sets the framework; the manager still runs live deal reviews and coaches on real accounts. The simulation's job is narrower: supply the volume of repetition that neither can supply on its own.
Frequently asked questions
Does this replace live coaching or manager-led role-play?
No. It sits before those sessions, not instead of them. A rep who has already run a scenario several times on their own arrives at a coaching session or a live deal review with a first attempt already behind them, so the manager's limited time goes toward refinement rather than a cold run.
Isn't this just a chatbot doing role-play?
The distinction is in what the simulation is built to hold onto across a conversation: a specific persona with a defined position, objections that don't move just because the rep asks nicely, and pressure that escalates or eases based on how the conversation is actually going — not a scripted back-and-forth that resets each time. A session is scored against an explicit rubric tied to the skill it's meant to test, not a single generic tone or confidence score.
How do we know it's actually working, and not just generating completion numbers?
Sessions completed is an activity metric, not an outcome one. The more useful measures are the ones a sales leader already tracks: rubric scores on a named skill improving across repeated attempts, and whether that shows up downstream in real pipeline movement — win rate on deals with a negotiation stage, or how often a stalled deal actually gets unstuck — over a full quarter rather than a single training cycle.
Does this work for negotiation specifically, or just discovery calls?
Negotiation is where the pressure and persona modelling matter most, which is why it's built out as its own track — multiple negotiation personas and scenarios, evaluated on both what the rep achieved and how they got there, structured as a certification ladder a rep can progress through rather than a one-time pass or fail.
Conclusion
None of this is a case against the workshop. It's a case for giving reps somewhere to fail cheaply and often between the workshop and the call that counts — the nets, not the match. For an L&D head or VP Sales weighing where deliberate practice fits into an existing training calendar, the place to start is naming the two or three conversations reps are actually losing deals over, and building the net around those first.