Picture this. You have 12 minutes to come up with a cover story that explains why you were caught going through secret papers. Then you’re led into a dark room with a spotlight in your eyes. The people asking the questions start out friendly. Then they jump on every slip.
That was a real selection exercise, run by American intelligence during the Second World War (WKU).
Twelve minutes to prepare, then questions that keep getting harder. If you’ve ever run a sales role play with a tough buyer and a few curveball objections, you’ve seen a gentler version of the same idea.
Everyone wants to know which traits to look for in a salesperson. The spies’ real lesson wasn’t about what to look for. It was about how to look.
Smart, well trained… and failing
In 1943, the Office of Strategic Services had a hiring problem. The OSS was the forerunner of the CIA. A field report complained that its recruits were smart and well trained but short on common sense. Several had to be pulled out of the field (ClearanceJobs).
If you’ve hired a sales leader who interviewed beautifully and then fell apart in the first six months, that probably sounds familiar. Smart, polished, great résumé… and not right for the job.
The OSS didn’t respond by hunting for a better kind of person. It changed the way it picked people.
The method they built
The fix was Station S, a country estate in Virginia. Psychologists put recruits through three and a half days of exercises, tests, a long interview and constant observation (ClearanceJobs). By the end of the war, 5,391 candidates had gone through Station S and a one-day version called Station W (CIA history).
There was no single gift they were looking for. Candidates were rated on a short list of qualities tied to the work, things like motivation, energy and initiative, effective intelligence, getting along with others, leadership, discretion, and the ability to observe and report (CIA history).
Two design rules did most of the work.
First, every quality was measured more than once, by different methods. Second, each exercise was built to look as much like the real job as they could make it (CIA history).
Read those two rules again and think about how most sales hires get made. One or two friendly conversations. A gut feeling. Maybe a reference call that confirms what we already decided. We check each quality once, if at all, and almost never in a situation that looks like the actual job.
One note: some of what the OSS did wouldn’t fly today. The interrogations, the sabotage exercises and the evening of drinks meant to crack cover stories were wartime methods. One psychology faculty page points out that the sabotage exercise wouldn’t meet today’s ethical standards (WKU). Nobody needs a spotlight to hire a territory rep. The design rules are the part worth keeping.
Where the method went
This isn’t just a war story. Station S became the model for the modern assessment centre. AT&T picked it up in the 1950s, and companies, police forces and governments still use versions of it today (ClearanceJobs).
And the research has mostly backed the spies up. A large 2022 review of hiring methods found that structured interviews (the same questions, scored the same way, for every candidate) were the strongest single predictor of job performance. Job-specific methods, such as work samples, ranked near the top too (Sackett et al., 2022). That’s research showing what works on average, not a promise about any one hire. But it points the same way Station S did. Test the real work, and test it the same way every time.
What didn’t transfer
Here’s where I have to be honest.
There’s no magic spy trait that predicts sales results. I couldn’t find any public study showing that the qualities used to pick spies also predict who sells more. The method transfers. The traits don’t.
The closest link I found is grit. A short grit questionnaire predicted which soldiers finished an Army Special Operations selection course, and which of 442 sales reps kept their jobs (Eskreis-Winkler et al., 2014). Interesting… but it was about staying, not selling more. And a later review of 88 samples, covering about 67,000 people, found grit is only moderately linked to performance and retention, and overlaps heavily with plain old conscientiousness. The part about keeping up effort over time carried most of the signal (Kaufman’s summary of Credé et al.; PubMed).
So don’t go looking for a grit score. Look for a track record of finishing long, hard goals, and ask references to confirm it.
The part nobody copies
My favourite part of the OSS story isn’t the dark room. It’s what happened after the war.
The staff graded their own work. They counted every failure in the field as a failure of their assessment, not just of the agent. They even admitted their early ratings ran high, because the first classes impressed them (ClearanceJobs, October 2026).
They also ran into a problem every owner will recognize. They often couldn’t get proper job descriptions, and agents were regularly sent to do different jobs than the ones they’d been assessed for (WKU). If you change the job after the hire, the mismatch isn’t all on the hire.
Most companies never do this kind of grading. A sales hire doesn’t work out, and we blame the person or the market. We rarely go back and ask what we predicted, why, and where our process got it wrong.
Three things to do on your next sales hire
Test the real work. Before the offer, give your finalists a realistic slice of the job. A discovery call with a tough buyer, or a deal that’s gone sideways. Score what they do in the exercise, not your read on their personality.
Check each quality at least twice. Pick the few things that really matter for the role, and check each one in more than one way. A structured interview question, the role play and a reference question about the same quality. One good conversation is one data point.
Write down your prediction, then check it. Before you hire, write down why you’re hiring this person and what you expect from them. Then pull out that note at 90 days and again at a year. It’s the cheapest way to find out whether your hiring process actually works… and almost nobody does it.
How we use it
This is the thinking we’ve built our own selection process around. Before we look at a single candidate, we sit down with the owners and leaders and write down what the job actually needs. Then we test each finalist on the real work, more than once, the same way every time, before you make the offer. If that sounds like the decision in front of you, start with our Final Candidate Review.
The spies didn’t find the perfect person. They built a better way to look. You can do the same on your next sales hire.
