we hired a bunch of actors

This is a guest post from my long-time friend and Sagan team member Brian Wilson:

I spent 13 years as a Counterintelligence/Human Intelligence Marine. For seven of those, I was an instructor and role player at the Marine Corps Counterintelligence Basic Course, the school that decides whether a Marine is ready to operate as a counterintelligence agent.

We had classroom instruction, SOPs, subject matter experts delivering lectures. All necessary. None sufficient.

The thing that actually told us whether a Marine was ready was the role play exercises. We built scenarios as close to real as possible. The Marine had to plan, prepare, and execute. In real time. Against a role player who wasn't going to make it easy.

We have a saying in the Marine Corps: "Rehearsal is where you make your money."

Role playing is the closest thing to game day without the consequences of game day. And most organizations aren't doing it.

Scaling It With Actors

When I brought this approach to Sagan, we hit an immediate constraint. We have close to 100 employees and hundreds of hours of role-play training to conduct. That volume isn't getting absorbed by already-overworked managers on top of their day jobs.

So we hired actors. Thirty bucks an hour. They follow a scenario brief, play a character, and give our team the reps they need without pulling managers off the floor to do it. It's one of the highest-ROI training decisions we've made.

Knowledge Is Not Performance

Your best people can explain the company's value proposition. They can tell you how a customer call should go. They can recite the process.

Then a customer says something unexpected, and they default to what they've always done. Not what they know they should do.

We built a role play replicating a first interaction with a new customer. Cooperative, friendly, lowest difficulty setting. We ran it with our most experienced customer-facing employees.

The exercise surfaced a clear pattern. There were specific moments where our team's responses, while natural and well-intentioned, weren't landing. They knew it. We knew it. The gap wasn't knowledge. It was how that knowledge showed up under the pressure of a live conversation.

That gap is invisible until you create the conditions to see it. Lectures, SOPs, and shadowing don't create those conditions. Role playing does.

Role Plays Stress-Test Your Systems, Not Just Your People

During our exercises, one team member used language we initially flagged as a coaching point. Then we went back and checked. The training document literally told her to say it that way. She executed perfectly. The training was wrong.

That error had been sitting in the document for months. No one caught it in review, in a team meeting, or reading the SOP. It took someone saying the words out loud to another human being to surface it.

When a pattern shows up across multiple experienced people, the first question isn't "what's wrong with them?" It's "what's wrong with our system?"

What We Did With the Data

Once we saw the pattern, we moved on three things.

We updated the training documents. The role plays showed us exactly where materials were unclear, outdated, or wrong.

We designed a targeted group session built around the specific moments that tripped people up. Not a generic refresher. Real examples from the exercises, breakout practice, reps on the exact skills that needed sharpening.

Then we run the role play again. Not as a test. As verification. Did the fixes actually close the gap?

That cycle is the whole point. Role play. See what happens. Fix the systems. Train the team. Role play again.

The Debrief Is Where the Learning Happens

The role play itself is just data collection. The value is in the debrief. Three steps.

The trainee self-assesses. "How do you think that went?" This matters because most people are either too hard or too easy on themselves. Calibrating self-assessment against reality is part of the development.

The role player shares what it felt like from the other side. Not "here's what you did wrong" but "here's what I experienced as the customer." When did I feel rushed? When did I feel heard? No SOP review or call audit replicates this.

The observer walks through the scorecard. Specific, observable behaviors. Did this happen or didn't it? GO or NO-GO.

You can't debrief a lecture. You can only debrief a performance.

Live Is the Hardest Setting

A live conversation is the most demanding thing you can ask a customer-facing employee to do. No backspace. No time to draft and revise. The customer says something, and you respond right now.

That's why you train there first.

Once someone can handle a live conversation, once they've seen how their words land in real time, every other channel gets easier. Writing an email? You've already tested the language out loud. Drafting a follow-up? You're not guessing. You've said it live, adjusted based on feedback, and now you're just putting what works into writing.

Train the hardest version of the skill. Everything else follows.

Start Simple

You don't need a military-grade program. You need four things.

• A scenario. Take a real customer interaction and turn it into a repeatable exercise. Write down the situation, the customer's mindset, and two or three moments you want to test.

• A scorecard. Five to ten observable behaviors. Binary checkpoints. Did this happen or didn't it?

• A debrief structure. Trainee self-assesses, role player shares the experience, observer reviews the scorecard. Ten minutes. Don't skip it.

• Reps. Run it. Fix what needs fixing, in the people and in the system. Run it again.

The first time you do this, you will learn something about your team that surprises you. Not because your team is bad, but because the gap between knowing and doing is universal. The only way to close it is practice.

If your team is customer-facing and you're not running role plays, you're hoping they perform under pressure instead of knowing they will.

Hope is not a training plan.

Yallah Habibi,

Jon