Knockout questions: which screening questions actually screen people out
Knockout questions from real data: across 14,598 answered screening questions, 'are you willing' questions almost never screen anyone out, while 'do you currently hold' questions catch more than a third.
· 6 min read · by the PhoneScreen.AI team
Every screening guide tells you to lead with knockout questions: the yes-or-no requirements that end a candidacy on the spot. Valid license? Can you work weekends? Is the pay rate okay? Ask those first, the advice goes, and you save everyone time.
The advice is right about the idea and quiet about the results. Most knockout questions knock out almost nobody. A few do real work. The difference comes down to how the question is worded, and we can see it in our data.
PhoneScreen rates every answer in every screen on its own, from 0 to 100, with a line of reasoning. As of today that is 14,598 answered questions from 2,352 completed screens across 82 hiring teams, mostly hourly roles: production and warehouse leads, caregivers, drivers, and assistants. About a third of those questions are yes-or-no requirements. The rest are open-ended.
For this post, an answer “screens someone out” if it was rated below 40. On a yes-or-no question that almost always means the candidate said no. We left out questions the candidate never reached, so a dropped call does not count as a no.
Most yes-or-no questions do not screen anyone out
Across all yes-or-no questions, 10% of answers were rated below 40. Open-ended questions came in at 34%.
So the questions built to filter candidates filter far less than the ones built to describe them. If you picture knockout questions as a gate that turns away a big chunk of the pool, our data says otherwise: nine in ten candidates walk through.
That does not make knockouts useless. It means you should know which ones are doing the work. Grouping the yes-or-no questions by what they ask about:
| Yes-or-no question about… | Answers | Teams | Rated below 40 |
|---|---|---|---|
| A credential they hold now (license, card, certification) | 241 | 19 | 36% |
| Schedule or availability | 448 | 12 | 15% |
| Everything else | 2,850 | 48 | 8% |
| Accepting the pay rate | 673 | 14 | 4% |
That is a ninefold spread, from 36% down to 4%.
“Are you willing” questions get a yes
The weakest knockout in our data is the pay question. “This role pays X. Does that work for you?” Only 4% of answers were rated below 40.
That makes sense once you think about who is answering. The pay was almost certainly in the job posting. Anyone who found it too low did not apply. The people left on the phone already accepted it when they clicked apply, and nobody talks themselves out of a job on the first call.
The same goes for most “are you willing to…” and “are you comfortable with…” questions. Willing to travel, comfortable with a fast pace, willing to learn new systems. Candidates on a screening call say yes to almost anything they can say yes to, and that is not lying. They are willing, in the abstract. The question does not ask them anything they could get wrong.
None of this means you should drop the pay question. Confirming pay on the call heads off a surprise at the offer stage. Just do not count it as a filter. It is a confirmation.
“Do you currently hold” questions work
The strongest knockout is the opposite kind of question: one about a fact the candidate either has or does not. A valid driver’s license, a current certification, a DOT medical card. 36% of those answers were rated below 40, across 19 teams.
The wording matters. “Are you able to get a forklift certification?” invites a yes. “Do you have a current forklift certification?” gets the truth, because the candidate knows you can ask to see it.
A credential is also something the rest of the screen cannot pick up. About a quarter of the candidates who failed a credential question still scored 60 or more overall. They sounded experienced and answered the open-ended questions well. Without the knockout, they would have looked like a strong candidate for a job they cannot legally start.
Schedule questions catch people the rest of the screen likes
Schedule and availability questions screened out 15% of candidates, which sounds modest. What happens to those candidates is the interesting part.
Nearly half of the candidates who failed a schedule question (48%) still scored 60 or more on the screen overall. These are good candidates. They have the experience, they answer well, and they cannot work Saturdays.
The open-ended questions measure whether someone can do the job. They do not measure whether the job fits their life. A schedule knockout is the one question that catches a strong candidate who would turn down the offer, or take it and quit in week three.
The catch is that schedule questions only work when they are specific. “Can you keep a flexible schedule?” is a willingness question in disguise, and it gets the same easy yes as the pay question. “This role is Tuesday to Saturday, 7 a.m. to 3:30 p.m. Does that work?” gives the candidate something real to say no to.
What our data does not tell you
Some knockout topics show up in fewer than ten teams’ question lists, so we are not drawing conclusions about them: reliable transportation, relocation, background and drug checks, and lifting requirements. For those, the numbers move too much with one team’s roles to say anything general.
Two more limits. These are mostly hourly and high-volume roles, so a director-level search will look different. And a low rating is not a rejection. PhoneScreen does not turn anyone down; it shows the hiring manager each answer and its rating, and they decide.
How to write knockout questions that work
Ask about what they have, not what they would do. “Do you have a current food handler card?” beats “Would you be willing to get a food handler card?” If getting the credential after hire is fine, it is not a knockout. Move it to the interview.
Name the actual schedule. Put the days and hours in the question. “Flexible” and “as needed” get a yes from everyone.
State the pay, then confirm it. Put the rate in the posting and confirm it on the call, but expect almost everyone to say yes. If nearly every candidate is passing your pay question, it is working as a confirmation, which is all it can do.
Keep the knockouts to two or three. Our phone screen length data shows each question adds about 40 seconds, and most of the value is in the few that separate people. A knockout nobody fails costs every candidate time and tells you nothing.
Keep them about the job. A knockout is a requirement of the role, asked the same way of everyone. Never ask about age, health, family plans, religion, or national origin. For physical requirements, ask whether the candidate can do the task, such as lifting 50 pounds regularly, with or without reasonable accommodation. We cover why the same questions for every candidate matters in more depth.
Spend the rest of the screen on open-ended questions. Knockouts tell you who cannot take the job. They do not tell you who will be good at it. A question like “Walk me through your last role” or “Tell me about a shift that went sideways” is where the spread between a 50 and an 80 shows up. How to read a scorecard explains what to look for in those answers.
Check your own list
Pull up the screening questions on one of your open roles and sort the yes-or-no ones into two piles: questions about something the candidate has or can do on your schedule, and questions about their willingness. The first pile is your knockout list. The second pile is mostly confirmations, and some can go.
If you want to see how often your knockouts actually screen people out, set the role up on PhoneScreen. Every answer gets its own rating, and the first ten screens are free.
Curious what the candidate hears?
Take the three-question sample screen for a paper sales job at Dunder Mifflin. Your scorecard lands in your inbox.