Client Profile

Growth-stage technology company | 45 employees | Functional leaders hiring specialized roles

What they believed

Leadership believed they had a volume problem. Every specialized role generated more material than anyone had time to read - resumes, interview transcripts, scattered notes from three or four functional leaders who had each sat with the candidate separately. Their stated ask was a faster way through the same pile. They were explicit that they did not want to buy another platform.

What the analysis showed

We built a custom evaluation tool for their environment. It took the complete record for each candidate - resume, interview transcripts, interviewer notes - and evaluated it against the competencies the role required and against the company's own core values framework, including the behavioral signals attached to each value. It produced a structured summary and directional scoring for each candidate, and a side-by-side view across the candidate slate.

The volume was not the problem. Two things surfaced that none of the leaders had seen: Leadership consistently described the ability to operate without structure as the single trait that separated a successful hire from an unsuccessful one in their environment. Across the complete interview record for the first several roles, no interviewer had asked a question capable of producing evidence of it. Each had assumed it was someone else's question.

Second, technical depth carried the heaviest weight in their evaluation, and every finalist cleared it - because the screen earlier in the process had already filtered for exactly that. The category that felt most decisive was doing almost no work in the actual decision, while the categories that genuinely separated candidates carried the least weight.

What changed

• Two questions covering the ownership trait were added to the interview guide and assigned to a named interviewer rather than left to the panel.
• The evaluation weighting was rebuilt so that the categories that differentiated candidates carried proportionate weight.
• Functional leaders began reviewing the structured comparison before the debrief rather than arriving with independent impressions and arguing from memory.

Where it stands
Median time from final interview to hiring decision fell from nine days to four. The tool was used on eleven of the next twelve specialized hires across seven months. On two of the first six roles, the analysis changed the order of the finalists.

This engagement is where the Talent Intelligence Engine came from. The tool built here was single-client and purpose-built; TIE is the productized version, with the scorecard calibration layer, the comparison report, and the executive debrief that this build made the case for. The pattern that justified building it is the second finding above - that a hiring team can be measuring something precisely and still not be measuring the thing that decides the hire.

Client Profile
Professional services firm | 55 employees | COO-led operations | Distributed teams

What they believed

Two strong finalists reached the final round for a senior client-facing role. All four interviewers ranked the same candidate as the top choice. Nobody dissented, and the decision was effectively made before the debrief started.

What TIE showed

The unanimity was not four assessments agreeing. Three of the four interviewers had cited the same info as their primary evidence - a story the candidate had told in three of his four conversations. One data point, counted three times. Outside it, the record held little evidence of the competency the panel believed it had confirmed. The second finalist had the weaker interview and the stronger record: less polished, far more specific, nearly all of it traceable to a particular engagement and outcome. The panel had been reading presentation quality as delivery quality.

What changed

• One additional structured conversation with both finalists, built around the competency the evidence was light on.
• The ranking changed. The firm hired the candidate all four had placed second.
• A new rule in the debrief: when two interviewers agree, find out where each of them got it. Same source means one opinion, not two.

Where it stands

Eleven months in, the hire is still there. Whether the other candidate would have worked out is unknowable, and we won't pretend otherwise. What is knowable: this firm was one debrief away from making a six-figure decision on a single story told three times, and they caught it - because TIE uncovered the need to walk back through the whole record and ask where each conclusion actually came from. That's what decision science actually buys you.

Client Profile

Design/build construction firm | 35 employees | Owner-operator led | Project-based growth

What they believed

The owner had already decided to promote his best lead field operator to superintendent. Eighteen years with the firm, the man everyone called when a detail was hard, and a crew that would follow him anywhere. It wasn't really a decision. It was a reward, and it was overdue.

What TIE showed

Measured against what the superintendent role actually had to accomplish, nearly every piece of evidence in his record was about work he had personally performed. There was almost nothing about getting work out of other people. And on the two things the job most often fails on - holding a subcontractor to a date that costs the sub money, and telling an owner bad news early - there was no evidence at all. Not weak evidence. Literally none. Nobody had ever asked. The firm's own record pointed the same direction. Three supers had come off the field in six years. The two who struggled went straight from the tools to running a job; the one who took to it had spent two years as an assistant super first, running work that wasn't his own. The difference between them wasn't talent. It was whether anyone had ever let them practice.

What changed

• He was promoted - to assistant superintendent on a smaller job, with the missing piece named out loud and a senior super assigned to him.
• Holding subs accountable and communicating bad news early became stated requirements for the role, with interview questions built to produce evidence of them.
• The owner stopped treating the job as something you earn by tenure and started treating it as something you get ready for.

Where it stands

It's been fourteen months, he's running his own job. Whether he'd have made it the other way, nobody can say. But the firm was a week from handing him a nine-million-dollar schedule on the strength of eighteen years of work he'd done with his own hands, and they caught it before the start date instead of in month four. He still got the job. It just came in the right order.