A Structured Task That Holds Up Under Questioning Reveals The Sales Talent An Interview Can't

The standard sales interview is built to be beaten. Candidates who sell for a living are, by definition, skilled at presenting the best version of themselves, which means the format most companies rely on rewards the polish that makes strong and weak hires hard to tell apart. The result is a hiring process that tests presentation when the job is about substance. The way to cut through the veneer is to rebuild the process around evidence a candidate must produce rather than claims that they can perform.

Engineering such a process is Scott Boocock, Vice President of Sales at the client service intelligence platform Kaizan. He's spent his career in enterprise sales leadership, including running Treasure Data's business across EMEA and India and strategic account work at Adobe, and is now building a sales team from the ground up. To do it, Boocock has replaced the conventional interview with a multi-stage process anchored by a written exercise, which has netted a series of successful hires. At a company still scaling its revenue engine, each of those hires comes with distinct stakes.

"Every person I hire is mission critical because they're responsible for generating revenue. You can't expect a company to make that kind of decision off the back of a couple of interviews and a walk through your CV," Boocock says. The task exists to surface the things a conversation can hide.

Why sellers are the hardest candidates to read

The problem Boocock is solving is specific to the role. "Salespeople are trained to put forth the best version of themselves," he says. He can spot an obviously good or obviously bad candidate, but the interview format collapses the distinction that actually matters, which is discerning how a strong presenter really works from how they present. So, after an initial chemistry call to check basic fit, he moves promising candidates straight to a written task they complete on their own time.

The exercise is structured in three parts, each targeting something a CV leaves untested. First, the candidate has to demonstrate that they understand what Kaizan does. "Tell me who our ICP is and what we're doing for them. What are the problems that we're solving using our technology?" Second, they build the raw materials of the job: a LinkedIn connection request and email to a given persona, plus an agenda for a first client meeting broken down by how they'd spend the time. Finally, they work a classic deal scenario, a champion whose CFO kills the budget at the last minute. "How do you tackle that? And the other bit, how do you stop it from happening in future deals? I'm looking for people who can work through a problem."

The interview becomes an interrogation of the work

The task isn't the test on its own. It becomes the spine of the first face-to-face interview, where Boocock pushes candidates on what they produced. "I'll say, 'Talk me through your thought process. That part doesn't resonate with me. This part, rewrite it live,'" he shares, describing how he works through the response in the room. Asking a candidate to reconstruct their reasoning, and to redo a weak section on the spot, is what exposes how they actually think rather than what they were able to prepare. That structure has repeatedly overturned his first impressions. Candidates who charmed him in the chemistry meeting have fallen down on the task, and people who didn't stand out on paper have earned a place because more of them showed up in the work.

It's also become an unexpected filter for AI-generated effort. Boocock tells candidates they're welcome to use AI for ideas but not to produce the task, and he can spot the difference. Some submit polished, clearly machine-made decks anyway. "One poor candidate was reading from the screen when she came in, and it was clearly the first time that she'd read it," he recalls. The gap between the artifact and the person is the signal.

Where AI sharpens human judgment

The same technology that trips up shortcut-takers is what Boocock uses to pressure-test himself. Every CV he receives runs through Claude against a clear brief for a first-pass shortlist, which he then gut-checks rather than takes on faith. Ahead of each first interview, he has Claude read the CV, the task, and the chemistry-call transcript and surface five things to dig into. "If it was left all to me, I might miss some of those. I might have just three that stand out," he says. The model catches inconsistencies he might overlook, like a deal whose numbers don't fit the logic of the account, a figure worth questioning, or a claim in the task that contradicts the CV.

Kaizan, Boocock's own company's product, runs underneath the process as the record. It captures and summarizes every interview, giving him a searchable account of what each candidate said and flagging any moments that don't add up. He also runs interview transcripts through Claude to build a scored leaderboard, with his rating against the model's. "I always take my score over Claude's, but if Claude's is way off, I try and figure out what I've missed," he says, noting that the model is working with the full transcript while he's working from the moments that stuck with him.

The fairness question, answered by the job itself

The obvious objection is the burden to candidates: unpaid work with no guarantee of an offer. Boocock meets it head-on. A large company can absorb a bad hire; a fast-growing one building its revenue engine cannot, which is why he considers the depth of the process proportionate to the stakes of sales hiring. He also points out that the task mirrors the job it's hiring for. "Customers are going to ask them to do exactly the same task to prove the value of Kaizen. That's just part of the job. I think salespeople, maybe more than other types of candidates, understand that." He also closes the loop by putting in some extra effort of his own: any candidate who asks for it receives feedback on their task, a walkthrough of what landed and where they fell short. "People have been pretty appreciative of that."

Other than successful hires, the evidence that Boocock's process is working is in the feedback. Recruiters warned him the process would repel candidates, but the candidates themselves have shared the opposite. "A lot of them have said that going through the task actually made them more aware of what we're doing, and it made them want the job more," he says. For a role where every hire moves the number, Boocock treats that self-selection as a feature. The people who lean into the work are the ones he wants at the top of the leaderboard, which is the hardest place in sales hiring to see clearly.