Paid Tests and the Specificity Interview
Replace interview claims with paid sample work and a fixed ladder of number-level follow-up questions.
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 84%
Sam Parr names two changes that improved his hiring. The first is the paid test: he spends three or four hours writing a brief, then pays the candidate to spend a day on it, from around two hundred dollars up to two thousand depending on the role. He would rather see the work than hear someone describe it, and it saved him significant time overall. The second came from watching a friend who had invested in one of his companies run three interviews for a make-or-break chief marketing officer hire. Sam interviewed candidates first, his friend interviewed them second, and Sam repeatedly liked people his friend rejected. The difference was specificity. Where Sam asked how they would rate themselves and asked for a story, his friend asked what the revenue was when they started, how it grew over the first eighteen months, and which three things they did that most impacted that. Not knowing the number was disqualifying on its own. He then asked the candidate to screen share and walk through their thinking out loud. Underneath both changes is Sam's framing that a leader's job is decision-making, over and over, and that decisions ten to twenty percent better each month compound into a very different outcome.
Origin
The specificity ladder came from Sam deliberately putting a more experienced friend on his CMO interviews and listening rather than leading, so he could watch how a master does it and raise his own game.
Core principles
- 01Words equal lies, work does not
- 02Pay for the test, do not extract free labour
- 03If they do not know the revenue number, they are out
- 04Ask for three specific wins, not one flattering story
- 05Watch someone better than you interview, then copy them
How to run it
- 1
Build the paid test brief
Invest three or four hours writing a document that sets out a real brief for the role. The quality of the brief determines whether the output is diagnostic.
- 2
Pay for a day of real work
Offer the candidate payment to spend a day on the brief, scaled to the role from roughly two hundred dollars to two thousand. Paying keeps it honest and gets you serious candidates.
Pro tip The point is to see what they do, not to hear them tell you they are good at work.
- 3
Open the interview with the baseline number
Ask what the revenue was when they joined their last role. If a marketing leader cannot say how revenue changed from day one to month twelve, end it there.
Watch out Do not accept a range or a shrug. Not knowing the number is itself the answer.
- 4
Ladder into the three impact moves
Once they give the number, ask what the three things were that they did that most impacted it. If they cannot name three wins across three years at the company, that is disqualifying.
- 5
Make them screen share and think out loud
Ask the candidate to share their screen and walk through how they actually approach the thing they just claimed to have done, talking out loud so you hear the reasoning rather than the summary.
- 6
Put a better interviewer in the room and listen
For make-or-break hires, bring in someone who has made great hires, let them lead the interview, and stay quiet. Sam did this deliberately with a founder friend to see where his own questioning was too shallow.
Pro tip Interview separately first, then compare verdicts. The disagreements are where you learn.
In the wild
Sam brought in a friend and investor to lead three interviews for a chief marketing officer hire he considered make or break. Sam would interview a candidate, like them, and his friend would come back and call them trash. Pressed for why, the answer was always that he had pressed for specifics and the candidate had none.
→ Sam adopted the specificity ladder, concluding there was a different level of detail he needed to be operating at in his own interviews.
Rather than a further round of conversation, Sam writes a brief over three or four hours and pays the candidate two hundred to two thousand dollars to spend a day producing real work against it.
→ He reports it saved him a ton of time in the end, because the output settles questions that interviews leave open.
Common mistakes
Asking candidates to rate themselves
Self-rating questions and 'tell me a story' prompts hand the candidate control of the evidence. Sam's earlier process asked exactly these and let weak candidates through.
Accepting one story as proof of a track record
A single well-told anecdote is easy to prepare. Demanding three specific wins with numbers attached is much harder to fake and reveals whether the person actually drove the outcome.
Running an unpaid test
Asking for a day of real work without paying for it filters out exactly the candidates with options, which inverts the purpose of the exercise.
Is it for you?
Best for
Founders hiring for roles where the output is observable, such as marketing, ops, content or design
Not ideal for
Roles where meaningful work cannot be sampled in a day, or candidates who cannot take paid outside work
From the transcript
“I would say two things have improved my hiring. Number one, paid tests instead of interviews.”
“I will spend the time I'll spend three or four hours creating a paid test.”
“he presses for specifics way more than I was previously doing”
“What were the three things that you did that most impacted that?”
From the episode
CEO Lessons: Bulls**t Mission Statements, Firing People & Saying ‘NO'
CEO Lessons