---
title: "How to reduce time to hire without hiring worse people"
description: "Most of your time to hire is waiting, not working. Here are the four queues that own the calendar, and which one to remove first."
url: https://zjcv.com/journal/how-to-reduce-time-to-hire/
canonical: https://zjcv.com/journal/how-to-reduce-time-to-hire/
author: "Zen Job CV Team"
published: 2026-06-10
updated: 2026-06-10
category: "Playbooks"
tags: ["time to hire", "hiring ops", "recruitment process", "hiring funnel", "interview scheduling"]
lang: en
---

# How to reduce time to hire without hiring worse people

> **TL;DR** Time to hire is a queueing problem: applications sitting unread, calendar coordination, feedback latency, and decisions waiting for a recurring meeting. Effort on a single hire is usually one or two working days; elapsed time is weeks. Measure elapsed days per stage, find the largest queue, and remove it, usually by pre-booking interview windows and screening in rolling batches. Do not buy speed by cutting structure: structured interviews are meaningfully more predictive of performance than unstructured ones.

Most of your time to hire is not work. It is waiting. Applications sit unread, candidates wait for a slot in someone's calendar, a hiring manager takes four days to send feedback, and a decision waits for a meeting that happens on Thursdays. Add up the hours anyone actually spends on a hire and it rarely exceeds a working day or two. Add up the elapsed time and it is weeks. Attacking the work makes almost no difference. Attacking the queues makes all of it.

That distinction matters more in a tight market. Eurostat put the euro area [job vacancy rate at 2.3% in the first quarter of 2026](https://ec.europa.eu/eurostat/web/products-euro-indicators/w/3-16062026-ap), with the Netherlands highest at 4.0% and Belgium at 3.4%. Where vacancy rates are that high, strong candidates are usually in more than one process, and the slowest queue in your pipeline is the one that loses them.

## Where the days actually go

Break a hire into elapsed time rather than effort and the shape is consistent across almost every team that measures it:

| Stage | What it costs in elapsed time | What causes the delay | Where the fix lives |
|---|---|---|---|
| Applications arrive, nobody reads them | Days to weeks | No owner, no scheduled slot, volume anxiety | First-pass process |
| First pass to shortlist | Hours of work, days of elapsed time | Screening is done in one heroic block nobody schedules | First-pass process |
| Shortlist to first interview | Often the single largest block | Calendar coordination across three people | Scheduling |
| Interview to feedback | One to five days per round | Feedback has no deadline and no format | Feedback discipline |
| Final decision | Days | Waits for a recurring meeting | Decision rights |
| Offer to acceptance | Days | Approval chains, comp sign-off | Pre-approved ranges |

Two of those six are about reading CVs. Four are about people waiting for other people. If you only ever optimise the screening step you will cap out quickly, which is why time to hire is a queueing problem wearing a recruitment costume.

## Why going faster usually makes hiring worse

The obvious lever is to spend less time on each candidate. It is also the one that reliably damages the outcome, because the parts of hiring that predict performance are the structured parts, and structure is exactly what gets dropped under time pressure.

The evidence here is old and unusually consistent. Schmidt and Hunter's synthesis of decades of selection research found structured interviews substantially more predictive of job performance than unstructured ones, with mean validities of roughly .51 against .38 ([Schmidt and Hunter, 1998](http://homepages.se.edu/cvonbergen/files/2013/11/The-Validity-and-Utility-of-Selection-Methods-in-Personnel-Psychology_Practical-and-Theoretical-Implications-of-85-Years-of-Research-Findings.pdf); the [2016 update by Schmidt, Oh and Shaffer](https://home.ubalt.edu/tmitch/645/session%204/Schmidt%20&%20Oh%20validity%20and%20util%20100%20yrs%20of%20research%20Wk%20PPR%202016.pdf) reaches the same conclusion on a larger base). A structured process is not slower because it is thorough. It is slower to set up once and faster to run every time after that, because nobody is inventing the questions, the criteria, or the scoring on the day.

The same holds at the screening end. A first pass done consistently is quick precisely because the criteria were decided in advance. A first pass done by intuition is slow, because every CV restarts the argument about what you are looking for.

So the goal is not a faster process. It is a process with fewer waits in it, run to the same standard every time.

## Queue one: the unread pile

This is the cheapest queue to fix and the most commonly ignored. Applications arrive continuously and get read in a panic, often a week or more after the role closed.

Three changes remove most of it:

**Give the first pass an owner and a slot.** Not "the team screens applications", but a named person with a recurring block. Work that belongs to everyone happens on nobody's calendar.

**Decide the criteria before the ad goes live, not after the applications arrive.** Must-haves and nice-to-haves, written down, in language a colleague could apply without asking you what you meant. This is the thirty minutes that pays back most across the whole process, and it is also what makes the first pass mechanical rather than agonising. The mechanics are in [how to screen hundreds of CVs](/journal/how-to-screen-hundreds-of-cvs/).

**Screen in rolling batches rather than one final sweep.** Reviewing on Monday and Thursday while the role is open means the shortlist exists the day the ad closes, instead of starting then. It also means the strongest early applicants hear from you while they are still available.

A screening tool helps here only in a specific way: it collapses the first pass from a block of work someone has to schedule into something that has already happened by the time they sit down. Zen Job CV produces a ranked list with the reasons attached, and the recruiter's job becomes reviewing a shortlist and a near-miss pile rather than reading two hundred attachments from scratch. The reading still happens. It happens on twenty people instead of two hundred.

## Queue two: scheduling

For most teams this is the largest single block of elapsed time, and almost none of it is work.

The pattern is always the same. A recruiter emails three interviewers for availability, two reply within a day, one replies on day three, the proposed slots have expired by then, and the loop restarts. Four calendar days disappear into an exchange that contained perhaps fifteen minutes of actual decision.

What works:

- **Pre-book the slots.** When a role opens, interviewers hold two recurring windows a week for the duration. Unused slots get released on the morning. This inverts the problem: instead of finding time after you find a candidate, the time already exists.
- **Let candidates self-book into those windows.** Any scheduling tool does this. The saving is not the recruiter's clicks, it is removing an entire round of email latency.
- **Cap the number of interviewers.** Every additional person in a loop multiplies the coordination cost and adds very little predictive value once the process is structured.

## Queue three: feedback latency

Interview feedback that arrives four days later costs you twice: it delays the next step, and it is worse feedback, because the interviewer is reconstructing an impression rather than reporting one.

Two rules fix most of it. Feedback is written within the hour, before the interviewer opens their next tab, and it is written into a scorecard rather than a free-text box. A scorecard takes less time to complete than a paragraph takes to compose, which is why the deadline is realistic rather than aspirational. It also makes the decision downstream faster, because three scorecards can be compared and three paragraphs of prose cannot.

The same logic runs backwards into screening. A shortlist where every candidate carries the criteria they met and missed is a shortlist a hiring manager can act on immediately. A shortlist of ten names with no reasons attached generates a meeting, and the meeting is the delay. This is the operational argument for [readable, explainable filter logic](/journal/plain-language-filters-vs-boolean-search/), quite separate from the fairness argument for it.

## Queue four: the decision

Decisions that wait for a recurring meeting inherit the meeting's cadence. If hiring decisions happen at Thursday's team meeting, your average decision delay is three days, and it is three days regardless of how quickly everyone did their part.

The fix is to name a decision-maker rather than a decision meeting. One person, agreed at the start of the role, who can say yes without convening anybody. Everyone else advises. This tends to meet resistance on the grounds of consensus, but consensus was never actually happening; what happens in the meeting is that the loudest opinion wins slowly instead of quickly.

Offer approval belongs in the same category. Agree the compensation range before the role opens and the offer needs no approval loop at all.

## What you should not cut

Some slowness is the process working. Cutting it buys days and costs hires.

Keep the structure: the same questions, the same criteria, the same scorecard for every candidate. That is what the validity research is measuring, and it is the difference between a fast process and a fast guess.

The candidate-facing symptom of these queues is [recruiters who seem slow to respond](/journal/why-recruiters-slow-to-respond/), and the interview stage in particular needs [its own connected tracking](/journal/how-to-track-interviews/) so nobody stalls between rounds. Keep closing the loop with everyone you did not hire. It buys no time to skip it, and it is where most of the reputational damage in a hiring process happens; the mechanics are in [how to reject candidates well](/journal/how-to-reject-candidates-well/).

Keep the near-miss review. A first pass that silently deletes everyone below a threshold is faster and also both worse and legally riskier, because a rejection with no human involvement is exactly the situation Article 22 of the GDPR restricts. The retention and automated-decision rules are set out in [the GDPR rules for CV screening](/journal/gdpr-rules-for-cv-screening/).

Keep the blind first pass if you run one. It costs nothing in elapsed time and it changes who reaches the shortlist, which is the point of the exercise rather than a nice-to-have. The evidence behind that is in [bias-aware blind CV screening](/journal/bias-aware-blind-cv-screening/).

## A worked example

One Product Designer role, 214 applications, a team of three.

Before: the ad closes on a Friday. Nobody reads anything for nine days. A recruiter spends most of a Wednesday on the pile and produces eight names with no notes. The hiring manager wants a call to discuss them, which happens the following Tuesday. Interviews are then scheduled by email over four days. Two interviewers send feedback the same week, one sends it the week after. The decision waits for the Thursday meeting. Elapsed time from close to offer is somewhere near five weeks, of which perhaps eleven hours was work.

After: criteria are written before the ad goes live. Screening runs twice a week while the role is open, so a ranked shortlist with reasons and a near-miss list exists on the Friday the ad closes. Interview windows were reserved when the role opened and candidates self-book into them. Scorecards are due within the hour. One named person decides. Elapsed time drops to somewhere near two weeks, and not one stage of the process was made less rigorous. The work was the same eleven hours. The waiting was mostly deleted.

Stage by stage, the change is entirely in the waiting:

| Stage | Before (improvised) | After (queues removed) |
|---|---|---|
| Applications sit unread | Nine days | Zero, screened in rolling batches |
| First pass to shortlist | Half a day of work, days elapsed | Ready the day the ad closes |
| Shortlist to first interview | Four or more days of email | Candidates self-book pre-held slots |
| Interview to feedback | Up to a week | Within the hour, on a scorecard |
| Final decision | Waits for the Thursday meeting | One named owner decides |


## Where to start with your next role

Measure one thing first: elapsed days per stage on your last three hires. Not effort, not cost, just the date each stage started and finished. Almost every team that does this is surprised by which stage is the largest, and it is very rarely the one they were about to optimise.

Then fix the biggest queue, once. Usually that is scheduling, occasionally it is the unread pile, and in slower organisations it is the decision. Fixing one queue properly beats improving all six a little, because a queue you have removed cannot come back next quarter.

Zen Job CV addresses one of those queues: it turns the first pass from scheduled work into a shortlist that is already waiting, with a reason next to every name and near-misses flagged rather than dropped. That is a real week for high-volume roles. It will do nothing at all for your calendar coordination, and it is worth being clear about which problem you are actually solving before you buy anything.

## Quick answers

**What is the fastest way to reduce time to hire?** Measure elapsed days per stage on your last three hires, find the largest queue, and remove it. For most teams the biggest block is interview scheduling rather than anything to do with reading applications, and pre-booked interview windows with candidate self-booking removes most of it in one change.

**Does faster hiring mean worse hiring?** It does if you cut structure. Decades of selection research find structured interviews meaningfully more predictive than unstructured ones, so dropping the questions and the scorecard under time pressure damages the outcome. Removing waiting time between stages costs you nothing in quality.

**How long should CV screening take?** The work is small once criteria are written down; the delay is almost always that nobody has scheduled it. Screening in rolling batches twice a week while the role is open means the shortlist exists on the day the ad closes rather than starting then.

**Should we cut interview rounds to move faster?** Cap the number of interviewers before you cut the number of rounds. Each extra person in a loop multiplies scheduling cost and adds little once the process is structured, whereas removing a whole stage often removes the signal you were relying on.

**When will a screening tool not help?** When your bottleneck is downstream. If applications are read within a day and the delay is calendars, feedback or approvals, a screening tool improves a stage that was never the constraint. Measure the queues first.

---

Source: https://zjcv.com/journal/how-to-reduce-time-to-hire/
Author: Zen Job CV Team