How to Screen Machine Operators and Production Associates: 9 Questions and What to Listen For Screening Interview Template

What kind of production work do you actually do: CNC machining on a mill, lathe, or Swiss, injection molding, stamping or press, extrusion or blow molding, filling and packaging, hand assembly, or a welding or fabrication cell? Which machines did you run last week, how many at once, roughly what was the cycle time, and were you setting them up yourself or tending a machine somebody else set?

What this assesses: This is the calibration question and it does more work than the rest of the list, because the title tells you nothing and the processes behind it do not transfer. CNC machining lives in thousandths: offsets, insert wear, surface finish, and a machinist who can set up from a print is a different seat at a different rate than an operator who loads, unloads, and gauges. Injection molding is a cycle and a mold: short shots, flash, sink, heats and water, regrind ratio, and a mold change that takes real hours. Stamping and press work is die changes, feed and misfeed, slug jams, and part protection, and it is where guards get bypassed because the jam is right there and the shift is behind. Extrusion and blow molding are continuous processes where the skill is startup scrap and catching a dimension that drifts slowly over hours rather than fails suddenly. Filling and packaging is a rate the machine sets, so the operator's value is uptime, clean changeovers between SKUs, and catching a bad date code or label before a pallet of it ships. Hand assembly is takt time and repeatability, often with torque specs and a workmanship standard. The how many at once detail is worth asking for on its own, because running four presses is a fundamentally different job from standing at one, and so is a 40 second cycle against a 12 minute one. The setup versus tending answer is the single biggest determinant of what the person is worth to you. Strong answers pick a lane, name specific equipment and materials, give a cycle time, and are honest at the edges, something like ran two presses start to finish including mold changes, only ever tended the packaging line. Be cautious with a candidate who lists every process at the same level, who says they can run any machine, or who cannot describe one part they made. Do not screen on years alone either. Two years in a shop where a lead taught them setups beats eight years loading the same fixture.

Walk me through a changeover you have done start to finish: what came off, what went on, what you set or adjusted, how you knew the first part was good, and about how long it took. Who signed off before you ran the order?

What this assesses: Changeovers are where operator skill actually lives and where your downtime actually goes, so this question separates the two seats cleanly. Strong answers give a sequence rather than a summary: lock the machine out, pull the tooling or die or mold, load and align the new one, set stops, offsets, heats, or water, run a first piece, measure it against the print, get it approved, then run. Listen for first-piece or first-article approval by name, including who signs it, whether that is quality, a lead, or the operator themselves against a checklist. An operator who runs a whole order without that step is how a thousand bad parts get made on a setup nobody verified. Credit anyone who gives a realistic duration and ties it to the work, since twenty minutes on a simple fixture and three hours on a press die or a mold that has to come up to temperature are both correct answers to different questions, and a candidate who claims a fifteen minute mold change has either done something unusual or is guessing. Credit the quiet markers of a shop that took setup seriously: staging tooling and material during the previous run so the machine is not waiting, separating the prep that can happen while the machine runs from the work that needs it stopped, and knowing how much startup scrap is normal versus when the scrap is telling you the setup is wrong. Also credit knowing what they are authorized to set and what belongs to a setup tech, because that boundary is a safety issue as much as a skill one. Be cautious with a candidate who says maintenance does all the changeovers or that they just run what is in the machine. That is not disqualifying, plenty of good tenders exist, but price the role accordingly and know you are buying a supervisor's time on every job change. Be more cautious with one who says they eyeball the first part.

How do you know the parts you are making are good? Walk me through reading a work order and a print, which gauges you have actually used (tape, calipers, micrometer, height gauge, bore gauge, go/no-go, thread gauge), and what you do when a dimension is still in tolerance but drifting toward the limit.

What this assesses: The drift half of this question is the real test, because it separates an operator who holds quality from one who only inspects it. Strong answers check a first piece against the print, re-check at set intervals or counts rather than when they feel like it, write the readings down, and treat a trend as information: adjust an offset, change an insert, verify the gauge itself, and tell the lead before parts go out of tolerance rather than after. The answer that should worry you is in tolerance is in tolerance, keep running, because that operator will make scrap right up to the moment somebody else catches it. If they have plotted points on a control chart, listen for whether they understand that a run of points on one side of the center line means something even when every part passes. That is the clearest single marker of a production worker who has been properly trained, and it is rarer than you would expect. On print literacy, strong answers can find a dimension and its tolerance, handle basic geometric tolerancing, and check the revision level, since running to a superseded revision is a quiet way to scrap an entire order while doing everything else right. They also read the work order for quantity, material, and which operation they are actually on. On gauges, listen for depth rather than a list: calipers are good to roughly a thousandth and a micrometer an order finer, and a candidate who mentions zeroing a mic, checking a calibration sticker, or treating a dropped caliper as suspect has worked somewhere that cared. Be cautious with a candidate who treats quality as the QC department's job, who admits they run until a supervisor stops them, or who cannot name a gauge they have personally read. Score the answer against one [interview scorecard](/glossary/interview-scorecard) so measurement skill and not interview polish decides it.

Mid-shift your parts start coming out wrong. Pick whichever fits your work: a dimension walking out of tolerance, flash or short shots, burrs, a bad seal, or a jam every few cycles. What do you check and in what order, how far do you go on your own, and when do you stop the machine and call maintenance or a setup tech?

What this assesses: The order is the answer. Strong candidates stop making bad parts before they start diagnosing, and they contain what has already been made: quarantine everything back to the last good check, because the first question your quality manager will ask is how many, and an operator who cannot answer that has turned a tooling problem into a sorting project. Then they work cheap and likely before touching a setting. On a CNC that usually means insert or tool wear, chip buildup, coolant, workholding, and whether the part moved, plus checking the gauge before chasing the machine. On a molding job it means material and dryer, barrel heats, cushion, mold temperature, vents, and how much regrind is in the blend. On a press it means die wear, lube, and the feed. The specifics matter less than the habit of starting at the likely cause rather than at a dial. Then listen for the escalation boundary, stated clearly: what they are trained and authorized to adjust, and what belongs to maintenance. There are two failure modes here and both cost you money. The operator who stands idle waiting for a supervisor on every hiccup bleeds throughput all shift. The one who opens a panel, defeats an interlock, or climbs into a machine they are not qualified to touch costs you an injury, and that is the worse of the two by a wide margin. Credit anyone who closes the loop: they write down what happened, code the downtime to a real cause, and tell the next shift what is acting up. Be cautious with a candidate whose whole answer is turning it up until it runs, who gives no containment step at all, or who has never stopped a machine, since every experienced operator has shut one down and had to justify it afterward.

Walk me through exactly what you do before you clear a jam or reach into a machine, and how you confirm it is actually safe. What is your PPE on this kind of equipment, and tell me about a hazard or near miss you reported and what happened after.

What this assesses: Clearing jams, cleaning, and servicing equipment is when production workers lose fingers and hands, so this question is worth more than the rest of the safety conversation combined. On lockout, a strong answer has four beats and the fourth is the one that matters: stop the machine, isolate the energy, apply their own lock and tag, and then verify zero energy by trying to start it. That try-out step is what separates somebody who has been trained from somebody who has watched. Credit stored energy specifically, because that is where people who follow the procedure still get hurt: air to bleed down, hydraulic pressure, loaded springs, a ram or platen that can drop under gravity, capacitors, and a barrel still hot enough to burn. On guarding, the answer you want is that guards and interlocks stay on, that they do not tape a switch or block a light curtain, and that a missing guard means stop and tell somebody rather than run it anyway and go faster. Ask what they do when the guard makes the job slower, because that is the honest version of the question and the answer tells you whether you are hiring a corner-cutter. PPE should be realistic to the process: safety glasses or a face shield, hearing protection, steel or composite toe, cut-resistant gloves where they belong and no gloves at all near rotating equipment, nothing loose, no jewelry, hair tied back. An operator who volunteers that you do not wear gloves at a lathe knows the floor. On the near miss, strong answers name a specific hazard they flagged and what changed afterward. Mature operators report things; a candidate who has never reported anything in ten years either has not been paying attention or works somewhere that punished it. Be cautious with anyone vague on lockout, anyone who describes reaching in while the machine is jogging, and anyone who has never stopped a job over anything.

What rate or standard were you held to, and how was it measured? Tell me about a shift where you fell behind, what caused it, and what you actually did. Where would you refuse to go faster?

What this assesses: Production is a standing tension between speed and scrap, and operators wash out in both directions. Strong answers treat a number as normal and name the one they lived with: pieces per hour, cycle time against standard, scrap or reject percentage, first pass yield, downtime minutes, or OEE if their plant tracked it. Then they describe the honest ways to go faster, which are almost all upstream of the cycle: staging material so the machine is never waiting, keeping tooling sharp instead of running it until the finish goes, doing setup prep during the previous run, and fixing the small repeated stoppages that nobody bothers to log and that quietly eat an hour a shift. Equally important is where they will not trade. Skipping quality checks, running through a known tolerance drift, bypassing a guard, and under-reporting scrap are all ways to make today's number look fine and hand the problem to somebody downstream. An operator who says plainly that they would rather tell the supervisor the rate is not achievable than bury scrap is giving you the right answer, and it is worth noticing whether their last plant made that possible or punished it. Be cautious with a candidate who claims they never fell behind, since nobody has not, and with one whose explanation is entirely other people: the previous shift, maintenance, the material, the schedule. Be cautious too with anyone who dismisses production metrics as management noise, because they will fight your scoreboard all year. And be mildly cautious with the opposite extreme, the operator who cannot tell you roughly how long a cycle should take, because a production seat has to clear both the quality bar and the rate bar.

What do you write down during a shift: production counts, scrap and the reason for it, downtime and the code, lot or heat numbers, quality checks, notes for the next shift? Have you worked under ISO, GMP, food safety, or any other traceability requirement?

What this assesses: Almost nobody screens for this, and it is where plants quietly lose the ability to fix anything. Your scrap numbers, your downtime analysis, and every root cause meeting you hold are only as good as what an operator wrote at three in the morning. Someone who logs forty minutes as machine issue has told you nothing; someone who codes it to a specific cause hands maintenance something to actually fix. Strong answers describe real records and treat them as part of running the machine rather than paperwork bolted on at the end: counts, scrap with a reason code, downtime with a cause, the quality checks they performed and when, and a handoff note saying what ran, what it ran at, what is acting up, and what the next shift should watch. If the work was regulated, traceability should come up without prompting. In food, pharma, medical, automotive, or aerospace that means recording the material lot that went into the run, date and lot coding on the output, logging a metal detector or checkweigher verification at intervals, and knowing that a failed check puts a hold on product back to the last good one. Credit GMP and food safety specifics, which are habits rather than opinions: gowning, hairnets and beard nets, no jewelry, allergen changeover and sanitation between runs, and glass and brittle plastic rules. Credit automotive and ISO experience the same way: a control plan, written operator instructions, and signing off that they followed them. Be cautious with a candidate who calls it busywork, who has only worked where a lead filled out the log for them, or who cannot name one thing they recorded. In a regulated plant that is a real training cost and an audit risk. In an unregulated one it still means your downtime data is fiction.

This role is [shift] on a rotating schedule with mandatory overtime when we are behind. What shifts can you commit to, have you worked overnights or a rotation before, and how would you get here for a start time before the buses run? Roughly how many days did you miss at your last job, and why?

What this assesses: Shift fit and transportation break more production hires than skill ever does, and this is the question where a good day-shift worker will self-select out if you let them. Strong answers give specific, realistic availability, have actually worked nights or a rotation and can say honestly what it did to their sleep and their family, treat occasional mandatory overtime as part of the job, and have a dependable way to reach an off-hours shift: their own vehicle, a ride they have relied on for months, or a home close enough to walk or bike. Give the transportation answer real weight rather than treating it as a formality. A five in the morning or eleven at night start sits outside most transit schedules, and a candidate depending on a bus that is not running then, or on a ride from somebody whose own shift is about to change, will be late or absent by week three. Nobody volunteers this, so ask it directly. On attendance, asking for a number gets you a far more useful answer than asking whether they are reliable. Credit a candidate who gives a real figure with a real reason, a surgery, a sick kid, a car that finally died, over one who claims a perfect record, which almost nobody holds across a year of shift work. Be cautious with a candidate who wants days only but applied to a night posting, who hedges on weekends, who reacts badly to mandatory overtime, or who cannot describe how they would physically get to work. That mismatch does not show up in the interview, it shows up as a call-off in week two. Put the real overtime expectation in writing, because candidates hear occasional no matter how you phrase it out loud, and use [asynchronous screening](/glossary/asynchronous-screening) so an operator who is on a press until 3:30 can answer at all.

The job pays [rate] on [shift] and means standing your whole shift on concrete, repetitive motion, lifting up to [X] pounds, and [heat / cold / noise / dust / cleanroom or GMP gowning], which you would do with or without a reasonable accommodation. You would need steel or composite toe boots on day one. We run a pre-employment drug screen and a background check. Have you worked a full shift in that kind of environment, what hourly rate are you looking for, and when could you start?

What this assesses: Most production hires that fail inside two weeks fail right here rather than at the machine, so put all of it in the screen and say the uncomfortable parts plainly. A full shift standing on concrete is its own kind of conditioning, and a press floor at ninety five degrees in July or a food plant held just above freezing is a job somebody should get the chance to decline now instead of quitting on day three. Ask the physical part exactly as written, as job demands with the accommodation language, and stay out of health history. Name the boots on day one and mean it, because an entry-level candidate between paychecks may not have a hundred dollars for boots before their first check clears, and that is a fixable barrier that otherwise reads to you as a no-show. If you offer a boot allowance or a voucher, this is where to say so. Ask for a rate rather than guessing at one. A CNC setup operator and an assembler do not price the same, and a plant paying assembly wages will not hold a setup hand no matter how well the screen went, so finding that out now saves you three weeks. Name the drug screen and the background check directly. Some of the [candidate drop-off](/glossary/candidate-drop-off) at this question is people selecting themselves out, which costs you nothing today and costs you a scheduled week once they are on the calendar, and if you hire where cannabis is legal, settle your policy before you post because it changes the pool you are fishing in. Then run every candidate through the same questions as a [structured interview](/glossary/structured-interview), so a setup operator and a packaging tender land on one rubric instead of being compared on who writes a friendlier answer. That consistency is the whole point of [high-volume recruiting](/glossary/high-volume-recruiting) when sixty applications come in against one opening.