.png){kind=avatar}
Most hiring teams know personality shouldn't drive hiring decisions. And yet it does—consistently, across organizations.
Recent research from Textio analyzed more than 10,000 documented interview assessments across nearly 4,000 candidates. People who received offers were 12x more likely to be described as having a "great personality," 5x more likely to be called "friendly," and 4x more likely to be described as having "great energy." More than a third of interviewers had commented on a candidate's personality before an offer was extended.
This isn't a story about interviewers being careless. Most interviewers are doing their best to run good interviews, and most talent teams have some version of a structured process to try to standardize rigorous assessments—a guide, a question list, a scorecard. The problem is: having questions in front of you doesn’t mean you’re able to run a high-quality interview. Because: it's hard to know in the moment whether a candidate's answer actually demonstrated the competency you were probing for. It's hard to know what follow-up to ask when an answer is vague or incomplete. It's easy to get pulled off-track by an interesting tangent and realize at the end that you never got to half your questions.
When interviews go this way, it compromises the debrief and decision process. Feedback ends up capturing how the conversation felt more than what the candidate is really capable of. When that pattern repeats across a panel, debriefs drift toward "I just really clicked with her" and away from useful evidence. Decisions get more difficult to make and more difficult to defend.
A question list isn't enough. Here's what it takes:
Upfront alignment on what you're evaluating. When intake is slow or incomplete, the interview plan reflects it: vague criteria, questions that don't assess the correct skills, interviewers coming to the conversation with different ideas of what a strong candidate looks like. Getting aligned on competencies before anyone talks to a candidate is what makes the rest of the interview process work (or not).
Interview questions that connect directly to intake. Once competencies are defined, the interview plan should follow directly—questions mapped to what you said you're evaluating, assigned to the right interviewers, ready to use. When questions aren't carefully constructed to evaluate specific competencies, the data you get back is weak or irrelevant—not the evidence you need to make a confident decision.
Real-time interviewer guidance and prompts. This is the part a structured interview guide alone can't solve. An interviewer can have the right questions in front of them and still miss a vague answer, lose track of time, or forget to probe when a candidate gives a partial response. An AI interview copilot keeps the interview on track in real time—prompting follow-ups when answers are incomplete, tracking which competencies have been covered, and capturing notes in real time as the conversation happens. Without that kind of in-the-moment support, even a well-designed guide produces inadequate data.
Evidence capture that happens during the interview, automatically. The gap between "what happened" and "what I remember happened" is where personality creep lives. When feedback is reconstructed from memory or incomplete notes hours later, it reflects an overall impression more than what was actually said. Evidence captured in real time—verbatim answers mapped to competencies as the conversation unfolds—is what makes the debrief concrete and the decision defensible.
The Textio data reflects what happens when evaluation is left to process alone—when the framework exists but the tools inside the interview aren't built to support it.
An AI structured interview platform like Lavalier is purpose-built to close that gap: smart automation from intake through debrief, and an AI interview copilot that helps every interviewer run a rigorous interview, every time. The result is a debrief built on evidence, a comparison grounded in competencies, and a hiring decision the whole team can stand behind.
Try it free on your next role and see what changes when the evidence is clear →
