Cases
What actually happened, in our customers' words
Four stories, told the same way: the situation, what we built, and what changed. Every story here is anonymous: customers share what happened and keep their name and company off this page.
Real estate acquisitions
A dead deal closed for $35,000 after six practice runs
$35,000
Created by one conversation, after the rep had already lost that same conversation six times in practice. A United States real estate acquisitions team.
The situation
A house was under contract at $270,000. The only buyer in hand would not go above $210,000. Someone had to call the seller and ask them to accept far less than they had agreed to, or the deal was dead.
What they did
The rep assigned to that call was, in his manager's own words, the weakest closer on the team. He ran the conversation six times against a simulated seller who refused, argued, and walked away.
What changed
The seller agreed to $175,000. The practice did not turn the rep into a different person. It let him lose the argument where it cost nothing, so he did not lose it where it cost $35,000.
In their words
"He went into the war prepared. On the simulation this guy was just being stubborn, so when he went to go deliver it, it was ten times easier."
How acquisitions teams run this practice →
The conversation behind this case is written up in full in motivated seller call practice, and the pushback the rep had to hold through is mapped in wholesaling seller objections.
Why most stories are anonymous
Training data shows where a sales team is weak, which is not something most companies want on a vendor website. Most customers share the story and keep the logo. Industry and role are named, the company is not. That applies to every story on this page, including the one above. We print a customer's name only when they have asked us in writing to do so.
What we will not claim
No revenue attribution, no invented percentages, no rounded-up numbers. Where a result is real but was never measured with a number, we describe it in words and leave it at that.
Case 01 · Pharmaceutical company
Preparing medical reps for physician visits
The situation
Field reps had to handle objections tied to specific products and specific physician specialties. Generic training could not cover that, and field coaching only reached a small share of the team.
What we built
Two AI physician personas and four visit scenarios, with objections and terminology taken from the client's own field reality, plus scoring criteria agreed with their training team.
What changed
More than 200 practice visits ran during the pilot, and the client extended it to the full medical rep team afterwards.
In their words
"The way feedback is delivered, I genuinely loved it. It's constructive, really detailed. And the fact that the rep tried to dodge a question, the trainer still flagged it at the end. That's exactly what I wanted."
Case 02 · Real estate and construction
Getting new reps productive faster
The situation
The L&D team needed to shorten onboarding and raise training quality at the same time. The blocker was practice: every rep needed repetitions, and there were not enough coaches to give them.
What we built
Scenarios for the conversations new reps struggle with most, with scoring set against the criteria the L&D team already used, so results could be read against their existing bar.
What changed
More than 400 practice sessions ran during the pilot, and participants rated it with an NPS of 85. Reps asked to keep using it, which is the part the team did not expect.
In their words
"Everyone's skeptical of any tool, any training. Whatever we assign, half the team immediately calls it a waste of time. And here they're saying 'awesome,' 'I'll keep using this.'"
Case 03 · Software company
Running their own sales framework without IT
The situation
The training team wanted to configure scenarios and prompts themselves, against their own sales framework. Waiting on a vendor or on internal IT for every product change was the thing they were trying to avoid.
What we built
Access to the no-code scenario builder with their framework as the scoring model, so their own team owns the scenarios rather than filing requests for them.
What changed
A scenario now takes about 15 minutes to build. The pilot launched in one week, and reps get up to speed on product changes in two to three days instead of months.
In their words
"We were looking for a solution where we could set up all the scenarios and prompts ourselves. We needed to move fast, and pichi.ai lets us do exactly that."
Method
How we measure
Every number on this page comes from one of the following. If a metric is not in this table, we do not report it, and we agree the metrics with you before a pilot starts rather than after.
| Metric | What it means | Where the number comes from |
|---|---|---|
| Dialogue score | How a single conversation performed against your criteria | Scored per session, with each score linked to the moment in the transcript that produced it |
| Skill breakdown | Structure, discovery, objection handling, next step, scored separately | Same session data, split by criterion so a weak skill is visible on its own |
| Progress | Whether a rep improves on the same scenario over time | Comparison of that rep's later sessions against their own earlier ones |
| Session volume | How much practice actually happened | Completed sessions counted per rep and per period from platform logs |
| Participant NPS | Whether the people using it would recommend it | Standard 0 to 10 survey sent to pilot participants at the end of the pilot |
| Readiness | Whether a rep clears the bar on a certification scenario | Pass or fail against a threshold you set before the pilot, not one we pick afterwards |
| Agreement with live calls | Whether practice scores match what happens on real calls | Your own call reviews or speech analytics on the same reps, compared against platform scores |
Two things worth saying plainly. Scores describe practice conversations, so calibrate them against your own call reviews before you make decisions about people. And we do not attribute revenue to training, because that link cannot be proven from this data.
How we decide what is scoreable in the first place is set out in call scoring rubric, and the six ways a human score drifts are in sales call scorecard. Both sit under sales coaching software for managers.
Behind the cases
The scenarios these teams actually ran
None of the stories above started with a curriculum. Each one started with a single call the team kept losing, turned into a scenario with a buyer who refuses. The call moments teams reach for first are listed on sales roleplay scenarios, and the frameworks their scoring is usually built on are on sales methodology roleplay practice.
FAQ
About these cases
Why are the companies not named?
Because they asked us not to name them. Training data is sensitive: it shows where a sales team is weak. Most customers are happy to share what happened and not who they are.
Are these pilot numbers or full rollout numbers?
Everything on this page comes from pilots, and each story says so. A pilot is a limited group over a few weeks, which is why the numbers are about adoption, volume, and setup speed rather than annual revenue.
Can we talk to one of these customers?
Sometimes. We ask the customer first, every time, and we do not share contacts without their agreement. Email us with your industry and we will say honestly whether a reference is available.
Why is there no revenue number here?
Because we cannot prove one. Revenue moves for many reasons at once, and claiming credit for it would be dishonest. We report what the platform actually measures, and we help you set up your own before and after comparison if you want a revenue read.
Want the version with your numbers in it?
Create an organization, run your own scenarios, and read the report on your own team instead of someone else's. Email us if you want help setting the scoring criteria first.