Screen-to-Hire Is the Number to Watch

The word "pipeline" gives people the wrong picture of hiring. A pipe carries whatever you put in the top out the bottom, and if you want more out, you push more in. Hiring works more like a row of rooms connected by doors of different widths. The flow through the whole thing is set by the narrowest door.

In most mid-sized hiring teams, I'd guess the narrowest door is the hiring manager interview. Sourcing, the ATS, and recruiter bandwidth get most of the attention, but they usually have room to spare. And because teams keep working on the stages that aren't the limit, a lot of effort goes into making the wrong parts faster.

Consider the arithmetic. Suppose a hiring manager can set aside four hours a week for interviews. That isn't a failing on their part. They have a real job, and the open role is one more thing on top of it. At 45 minutes a conversation, four hours comes to five interviews a week, and only if nothing else comes up.

Now suppose a recruiter is working two or three open roles for that manager and passing along a dozen or more screened candidates a week. Even after tough filtering, several people each week are waiting for time that doesn't exist. The queue fills faster than it empties. Someone who applied three weeks ago is still waiting for a 45-minute conversation while the manager sits in their fourth product review of the day. In agencies, fast-growing companies, and industries where people leave all the time, the gap is wider.

What makes this hard to fix is that nobody is responsible for it. Recruiters are measured on sourcing speed and coordinators on scheduling. Hiring managers turn up and do their best. No one owns the conversion from "entered the hiring manager's queue" to "got hired," so no one notices that the limit comes from how the process is set up, and that working harder at any one stage won't move it.

When people do notice the queue, the usual response is to add screening steps: a technical assessment, another phone round, a take-home assignment. The idea is that more filters produce a cleaner shortlist, and they do. But they don't widen the door. If the manager has four hours, they still see five people. Extra filtering decides which five, and it has no effect on how many.

So if you can't change the number of interviews, the thing to change is what fraction of them end in a hire. Call it screen-to-hire conversion: of the candidates who reach a hiring manager interview, how many are eventually hired. It tells you whether your screening agrees with what your hiring managers actually select for. My guess is that in many teams it's low, which means most hiring manager interviews end in a no.

Every one of those no's used an hour of the one resource in the process that can't be scaled or handed to someone else. That's why conversion matters so much more than speed. Suppose one in eight of the people a manager interviews gets hired, and better screening moved that to one in three. The manager's calendar would look exactly the same, and it would produce about two and a half times as many hires.

Why is conversion low in the first place? Mostly because the information screening produces is coarse. Resume keywords, job titles, and years of experience predict how someone will do only loosely. So the manager ends up doing the real qualifying in the interview itself, and spends scarce hours on sorting that should have happened earlier.

That's the part of the problem we built Asendia for, and it's why we chose voice. Asendia is voice AI that holds a live spoken screening conversation with each candidate, evenings and weekends included. Speech lets the conversation adapt to what the candidate says. It can ask a follow-up question and press on an answer that's vague, which a keyword parser can't do and an async video prompt full of rehearsed answers doesn't do. That's harder to build than a questionnaire, but what you learn is how someone thinks and talks about their work, which is different information from how well they formatted a PDF. A candidate who can describe in detail how they handled a situation in a past job is telling you more than one whose resume has the right words on it, and those are the people a manager's few hours should go to.

After Asendia screens a pipeline, the recruiter gets a ranked shortlist in the ATS, with a structured summary of each candidate's qualifications and excerpts from the conversation itself. Since the first conversations happen whenever candidates are free, the shortlist can be waiting by morning. The whole point is to make the handful of interviews a manager can do each week count for more. Once AI is doing that first conversation, some of the old recruiting numbers stop meaning much, and the post on recruitment KPIs in a post-AI world goes through which ones to drop and which to start tracking.

If you're about to spend money making your pipeline faster, I'd look up one number first. Of the candidates who reached a hiring manager interview last quarter, how many did you hire? If it's low, going faster will mostly put the wrong people in front of your manager sooner. The door stays the same width, and what you can change is who walks through it.

Ready to transform your hiring strategy? Schedule a Demo with our founders today!