How USB Payments Scaled Sales Training with AI Roleplay

USB Payments sells payment processing to small and mid-sized businesses across a distributed sales floor, the kind of environment where a new rep’s first thirty days often decide whether they stick around. Bill Pace, a sales and training manager with three decades in the field, had trained hundreds of reps the traditional way: shadowing calls, live roleplay sessions with a manager, and a long stretch of trial and error before a new hire touched a real prospect. When USB Payments set out to grow the sales team without growing the manager headcount at the same rate, the company turned to AI roleplay to carry the load that manual training used to carry alone.

A Payments Sales Team Outgrowing Manual Training

Before USB Payments adopted Yoodli, sales managers spent 20 hours a week on manual training. Half of a manager’s time during onboarding went to roleplay alone. New hires got roughly 8 hours of roleplay practice before their first customer call, and after that, ongoing practice dropped to about 1.5 hours a month. Managers were the bottleneck for every rep’s practice reps, and the math didn’t scale. Add five new hires in a quarter and you added five new demands on the same set of managers, each of whom already had a full book of coaching, forecasting, and pipeline work.

Payments sales carries its own pressure on top of that. Reps have to explain processing rates, contract terms, and integration questions with precision, because a fumbled answer on pricing or a vague response to a merchant’s objection costs a deal. A team that scales headcount faster than it scales practice ends up sending underprepared reps into calls that matter. That gap between hiring pace and training capacity is where most sales training programs break down, and it is rarely visible on a hiring plan until reps start missing quota in their first quarter.

Merchant services also has a longer objection list than most sales categories. A prospect might push back on processing fees, ask about contract length, compare a quote against three competitors, or want a plain answer on how a new terminal integrates with their point-of-sale system. Getting a rep fluent on all of that used to mean sitting them next to a manager for weeks and hoping enough of it stuck before the manager had to move on to the next hire.

Building an AI Roleplay Program Managers Could Actually Run

USB Payments used Yoodli to give every rep a place to run AI roleplays against realistic buyer personas, on their own time, without booking a manager’s calendar. Reps could rehearse a pricing objection, a competitor comparison, or a first discovery call as many times as they needed, and get feedback on pacing, filler words, and how they handled pushback before a manager ever heard the call. The onboarding use case is a common entry point for sales teams making this shift, since new-hire ramp is where the manager bottleneck shows up first and hurts the most.

According to the published case study, self-paced training hours per rep rose from 18 to 120 annually, a 600% increase, once reps had a way to practice without waiting on a manager. That volume of practice came from moving the bulk of repetition off manager calendars and onto a tool reps could use whenever they had ten minutes between calls.

What the Program Looked Like Day to Day

The mechanics were simple enough that reps adopted the habit without much prompting. A new hire preparing for their first cold call would run an AI roleplay against a buyer persona built to sound like the small business owners USB Payments actually sells to, skeptical about switching processors, focused on cost, and short on time. The rep practiced the opening, took the objection about fees, and got feedback on filler words, talk-to-listen ratio, and whether they answered the objection directly or talked around it. They could run the same scenario five times in twenty minutes, something no manager’s schedule could support.

Managers stayed in the loop, but their role shifted from running every rep’s first ten roleplays to spot-checking recordings and coaching on the calls that showed a real gap. A rep who handled pricing objections well in AI roleplay but stumbled on integration questions got a manager’s attention pointed at that specific weak spot instead of a generic run-through of the whole pitch. That targeting is part of why manager time stretched further without anyone working longer hours.

Faster Ramp, Better Meetings, More Practice

The results USB Payments published are specific, not rounded. Seller ramp time dropped 50%. Meetings booked rose 19% within 30 days. Meeting quality, measured by more in-person and qualified meetings, improved 20%. Average call length increased 20% as well, which the company attributed to reps holding conversations longer instead of losing them early to a stumbled answer. Sales leaders also reported saving more than 6 hours a week that had previously gone to manual roleplay sessions.

Bill Pace described the shift directly: a rep with no prior sales experience came in, trained using AI roleplay for two weeks, and won her first sales call. He called it a first in thirty years of training reps. Another sales leader at the company said the change reshaped daily habits on the floor, describing a team where practicing before a call became the default rather than something reps did only when a manager scheduled it.

What Changed for Managers, Not Just Reps

The headcount math is the part that gets missed in a story about faster ramp times. USB Payments didn’t hire more sales managers to support a growing rep count. Managers kept their existing coaching load, but the hours they’d spent running basic roleplay drills moved to AI roleplay instead, freeing them to spend their coaching time on the calls and deals that actually needed a human’s judgment. One sales leader put it simply: Yoodli made their time with reps more strategic, because the repetitive part of training no longer competed for the same hours.

That pattern shows up whenever a company tries to scale sales roleplay across a growing team without scaling manager headcount at the same rate. Most reps want to practice. The real constraint is manager hours, and removing it changes how much practice a rep can log before their skills get tested on a real call, along with how a manager’s day gets spent.

Why This Matters for Payments and Other High-Precision Sales Teams

Payments sales is a useful test case for scaling sales training because the margin for error is thin and the sales cycle moves fast. A rep needs to handle pricing objections and integration questions correctly on an early call, and a company adding reps every quarter needs a way to get each one to that standard without adding a manager for every few new hires. USB Payments solved that by decoupling practice volume from manager availability. Reps ran roleplays around their schedule, managers reviewed the sessions that needed their attention, and the ramp period shrank because reps arrived at their first real calls having already made their mistakes in practice.

Other companies have found the same constraint in different parts of the sales motion. Harness, 75% reduction in sales-training review time, cut the time managers spent reviewing recorded practice calls. Manager time is the scarcest resource in most training programs, and it’s usually the first thing to run out when a team grows.

USB Payments now treats AI roleplay as a permanent part of how new reps ramp and how existing reps keep their skills current. The team scaled its training capacity without scaling its manager headcount, and the metrics it published, a 50% cut in ramp time and a 19% increase in meetings booked within 30 days, reflect a sales floor where practice stopped depending on whoever had a free hour that week. Companies facing the same growth math in payments, financial services, or any sales org where precision on the first call matters can find a similar path by looking at where their own managers spend their training hours today.