Async Interviews: Running a Hiring Loop Without a Live Call

Eleven candidates went through a hiring loop with no live interview, and nine of them completed the whole process within six days. The previous loop took eighteen days on average, most of which was scheduling.

The role was a senior content strategist for a distributed company. The team had lost two preferred candidates in the previous year to slower competitors, both after the third round.

Where the eighteen days went

Scheduling four interviews across four time zones consumed an average of nine working days, because each slot required three people to agree. The interviews themselves took four hours of candidate time and eleven hours of team time.

The evaluation happened after the last call, in a meeting that lasted an hour and mostly consisted of people repeating what they had already written in notes.

The written replacement

The loop became three written stages. A portfolio review with four fixed questions. A paid work sample of about three hours, marked against a rubric published to the candidate in advance. A written conversation with two team members over 48 hours, where questions and answers were exchanged as documents.

Every stage had a deadline and a named reviewer. No stage required two calendars to align.

The work sample

The paid sample was the contested part. Paying for three hours of work was approved after an argument about cost, and the argument was settled by comparing the fee against the recruiter’s day rate for scheduling.

The rubric was published to candidates because the team could not find a reason to keep it secret. Nine of eleven said it changed how they approached the task, which the team treated as a positive sign rather than a leak.

What the format lost

Small talk disappeared, and with it some signal. Two hiring managers said they missed the ability to ask a follow-up question in real time, particularly when an answer sounded rehearsed. The written stage allowed follow-ups but they arrived hours later, which changed the pressure.

One candidate withdrew at the written conversation stage, saying the process felt like an exam. That is a real cost, and the team recorded it rather than explaining it away.

The evaluation

Every reviewer wrote a score and a reason before reading anyone else’s. Scores were collected by the recruiter and shared together, which removed the drift that happens when the first speaker sets the scale.

Agreement between reviewers rose compared with the previous loop, though the sample of eleven is too small to claim much beyond consistency.

What the six days contained

The fastest candidate finished in four days, the slowest in nine. The schedule was driven by candidate availability rather than by team calendars, which is the reverse of the old arrangement.

Two of the eleven dropped out voluntarily, one for the exam feeling and one for an offer elsewhere that arrived first. The second is the failure the redesign was meant to prevent, and it still happened.

Two hires later

Both hires came through the written loop. Their first-month output was compared with the previous two hires and was not measurably different, which the team takes as evidence that the live interview was not doing the work people believed it was.

What the written conversation looked like

The two team members sent six questions in one document and the candidate answered in the same document over two days. Follow-ups were allowed inside the document, quoted under the answer they referred to. The result reads like an interview and is easier to review afterwards, because the reasoning sits beside the question rather than in someone’s memory.

One candidate used the time to revise an early answer, which the format permits and a live call does not. The team treated the revision as evidence of care rather than of uncertainty.

Eighteen days once went into scheduling four conversations. Eleven candidates now complete a loop in six days, and nine of them finish without a single calendar invitation.