Manufacturing · Discovery Call
Discovery Call Questions for Manufacturing: A 25-Minute Playbook for Plant Managers, Maintenance Leaders and VPs of Operations
You've got 25 minutes with a Plant Manager who is going to be pulled out of this call the second line 3 goes down. He already knows your one-liner — that's why he took the meeting. If you open by telling him again, you've spent your credibility before minute two.
Here's what he's going to hand you first: "our uptime isn't where it needs to be" or "we're eating too much scrap." That's layer one and it's worth nothing. Underneath it is a mechanism — the downtime code says mechanical, the tech wrote three words on the work order, and nobody actually knows why the press failed. Underneath that is the cost: Saturday overtime, an expedite charge on a gearbox that's 22 weeks out of Germany, and the throughput miss he had to explain to corporate. And underneath that is the stake — the tier 1 customer's supplier quality engineer who is one escape away from putting his plant on controlled shipping level 1 with a third-party inspector on site at his cost.
This playbook gets you to layer three or four before minute 16, quantifies it in the numbers he already reports up (OEE on the constraint, unplanned downtime hours by asset, planned-versus-reactive work order ratio, customer PPM), maps how an AR actually gets funded in his shop, and lands a dated next step. It also tells you what to say when he says "capex is locked for the year" — which he will.
The discovery call script
Say it in your own words. The structure is the part that matters.
- 1
0:00–2:00 — Frame (do not re-pitch)
"Thanks for holding the time — I know second shift start is right about now, so if you get pulled I'll take whatever we get. When we spoke Tuesday you said your unplanned hours on line 3 have been creeping since the spring and you couldn't tell corporate why. That's the thing I want to dig into. Fair warning on how I run these: I'm going to ask questions for most of this call and show you basically nothing. I'd rather tell you at minute twenty that this isn't a fit for your plant than burn a demo slot on you. Still good for 25? And is there anything you want to make sure we hit, so I don't run us out of clock?"
- 2
2:00–6:00 — Layer 1 to Layer 2: get to the mechanism, not the symptom
Open the thread with one question. One. "Walk me through your constraint. Which asset does everything route through, and what's it doing on a bad week?" Then do not move on. Three follow-ups minimum on the same thread: - "Walk me through the last unplanned event on that asset. What actually happened — who found it, who called it in, how long from stop to running?" - "When the tech closed that work order, what did the downtime code say? And what did he write in the notes?" - "So how do you know it was mechanical? Who decided that?" The third one is the crowbar. Most plants can produce downtime hours by asset all day and cannot produce a cause. When he admits the code is a dropdown someone picks at 3am, you've reached layer two. If he's a Maintenance Manager or Reliability Engineering Manager, swap the thread: "What's your planned versus reactive work order split running at last quarter — and is that trending anywhere?" If he's a Quality Manager: "Walk me through your last escape. Where in the process did it get made, and where should it have been caught?"
- 3
Minute 6 — "So what is it you guys actually do?" (the trap)
Thirty seconds, tied to what he just said, then hand the ball straight back. "Short version — we put condition monitoring on the assets you can't afford to lose, so instead of a downtime code that says 'mechanical' after the fact, your millwright gets a named failure mode with lead time on it. Nobody on your floor enters data for that to work. But I'd be guessing whether that matters to you until I understand one thing: on that six-hour stop last month, was the part on the shelf?" If he pushes a second time, he genuinely needs orientation. Give a clean 60 seconds, then: "Can I go back to the thing you said about the gearbox lead time? That's the part I'm not clear on." He'll let you.
- 4
6:00–11:00 — Layer 3: the cost, and then "how do you know?"
"How many unplanned hours a month is that asset costing you right now — not plant-wide, just that one?" "And what's an hour of downtime on the constraint worth to you? You've probably got that number for the AR you wrote last year." Then the beat most reps skip: "How do you know? Is that a real cost-per-hour your controller signed off on, or a number the plant uses?" Keep going down the same thread: - "When you lose six hours on line 3, what does the recovery look like — Saturday crew, third shift, both?" - "Where's overtime running as a percent of direct labor right now?" - "How much of last quarter's MRO spend was expedite and air freight you wouldn't have paid with two weeks' notice?" - "Did OTIF move? Did anything ship late to the customer because of it?" If he can't answer "how do you know," that is a finding, not a dead end. It means the problem is invisible above him — and you're going to have to help him build the number before anyone funds anything.
- 5
11:00–16:00 — Layer 4: the stake (who is personally exposed)
Say these slowly and then stop talking. Count to three. He will fill the gap. - "Whose number does this land in when it misses — yours, or does it roll up to the VP of Operations?" - "Who's feeling it most on your staff right now?" - "What did you commit to corporate for this year on the constraint — is there an OEE target on it?" - "Has any of this reached the customer? Any escapes last quarter, any movement on your PPM?" - "If nothing changes and we're having this same conversation next November, what's the conversation you're having with your VP?" When the Plant Manager tells you his 30-year millwright retires in eighteen months and there's nobody behind him, or that the tier 1 customer's SQE has already been on site twice — that's layer four. Write his exact words down. That sentence is your business case.
- 6
16:00–20:00 — Qualify the path (never recite BANT)
"If you decided this was worth doing, what actually happens next in your world? Who gets pulled in — reliability, IT, corporate engineering?" "What's your signing authority before it has to go to an AR? Twenty-five, fifty?" "Where are you in the AR cycle right now — is next year's already submitted, or is it still being built?" "Is this a line item that exists — MRO, controllable spend — or does one have to get created?" "What's your hurdle rate on payback? What's the number it has to beat to survive against the new press and the roof?" Then the graveyard question, which is the most valuable one on this page: "Have you tried to fix this before? What happened to it?" And finally: "What's forcing a timeline — a customer audit, a program launch, the millwright's retirement date? Or is nothing forcing it?" If nothing is forcing it, you've had a pleasant conversation with a plant manager, not a deal.
- 7
20:00–23:00 — Targeted relevance (90 seconds, only what he raised)
No slide tour. One mechanism, his words: "So on the specific thing you described — the downtime code that says 'mechanical' and nothing else. What we'd do on that press is put vibration and thermal on the drive end, and instead of an alert going to a shared inbox, it opens a work order in your CMMS with a named owner and a failure mode. Your Reliability Engineering Manager sees 'outer race defect, estimated 9 to 14 days,' which is enough runway to get the bearing in and do it on a planned Saturday instead of a Tuesday afternoon at line rate. Read-only, no PLC connection, and it can sit on its own cellular gateway that never touches your plant network — because I'm guessing your IT/OT guy is going to ask that first. That's the whole thing. Does that shape match what you'd actually need, or is the harder part getting the spare on the shelf?"
- 8
23:00–25:00 — Playback, then a dated next step with names on it
Playback in his words, three sentences: "Let me make sure I've got this. Line 3 is your constraint, you're eating roughly 40 unplanned hours a month on it, and the reason you can't get ahead of it is that nobody knows why it fails — the code says mechanical and Dave writes three words. It's costing you Saturday crews, overtime around 14 percent, and you had one late shipment to the tier 1 in September. And the pressure's real because Dave retires in eighteen months and the SQE has already been on site twice this year. Did I miss anything, or get anything wrong?" Then prescribe: "Based on that, the useful next step isn't a demo. It's 45 minutes with you, your Maintenance Manager and whoever owns IT/OT security, where we do two things: pull the last twelve months of downtime hours on line 3 and the two assets behind it, and build the payback number you'd need for the AR. If it doesn't clear your hurdle rate, I'll say so on that call and stop calling you. I've got Tuesday at 7am before your production meeting, or Thursday at 2. Which one?" Send the invite before you hang up. Confirm out loud: who's on it, what you'll show, and what he brings — twelve months of downtime by asset, and his cost-per-downtime-hour if he has it.
- 9
What goes in the CRM
His words, verbatim. Not your summary. Good: "Line 3 is a '98 Bliss press, everything routes through it. Dave Kowalczyk retires June '26, nobody behind him, tribal knowledge is a laminated sheet taped to the machine. 40ish unplanned hrs/mo. Signing authority $50K, AR submitted Q3 for next year. Tried Banner sensors in 2022, alerts went to a shared maintenance inbox, dead in six weeks. SQE from [customer] on site twice — 'one more escape and we're in controlled shipping.'" Bad: "Prospect has reliability challenges and budget constraints." The first version writes your follow-up email, your business case and your exec deck. The second version writes nothing.
How the call actually sounds
Prospect on the left, the rep on the right.
Rep
When we talked last week you said the unplanned hours on line 3 have been climbing since spring and you couldn't tell corporate why. That's what I want to spend the time on. I'm going to ask a lot and show you almost nothing — if it's not a fit I'd rather say so at minute twenty. Still good for 25?
Buyer
Twenty-five, and I've got a supplier quality guy walking the floor at ten so I mean it. And I'll tell you up front — capex is locked. Whatever this is, it's a next-year conversation at best.
Rep
Noted, and I'm not going to pretend I can unlock it. Tell me about line 3 first. What is it, and what's it doing on a bad week?
Buyer
It's a 600-ton press from 1998. Everything routes through it — if it's down, four downstream cells are down inside two hours. Bad week is six, eight hours unplanned. Last month it was closer to forty across the month.
Rep
Walk me through the last one. What actually happened, from the stop to running again?
Buyer
Ram wouldn't cycle. Second shift called the on-call millwright in at eleven, he was there till five in the morning. Turned out to be the main drive bearing. We ran Saturday to catch up.
Rep
When that work order got closed out, what did the downtime code say?
Buyer
Mechanical. That's the code. Look, I know where you're going with this and I'll save you the trip — we're at 95 percent PM compliance. My CMMS is clean. We're not a run-to-failure shop.
Rep
I believe you, and most plants I walk aren't at 95. But those are two different metrics. If PM compliance is 95 and you're still eating forty unplanned hours on the constraint, that usually means the intervals are calendar-based — you're changing bearings that had life left and missing the ones that didn't. What's your MTBF trend on line 3 over the last twelve months?
Buyer
...Flat. Maybe worse. Which is the thing that bugs me, honestly. We're doing everything on the schedule and the number isn't moving.
Rep
What did the forty hours cost you last month, all in — the Saturday, the overtime, the expedite on the bearing?
Buyer
I can give you downtime hours by asset to the minute. Cost per hour I've got a plant number, but I don't love it. Corporate says twenty-two hundred an hour on the constraint. My controller and I have argued about whether that's real.
Rep
That argument is worth having before you write the AR, not during. Separate from the hours — did any of this reach the customer?
Buyer
One late shipment in September. And the reason there's an SQE on my floor at ten is we had an escape in August. We're not on controlled shipping but I'd call it a warning shot.
Rep
How many escapes is that for the quarter?
Buyer
Two. Which for that program is two too many.
Rep
Who's feeling that most right now?
Buyer
My quality manager wrote the 8D at nine o'clock at night and she's already carrying two open corrective actions. And honestly — the guy who actually knows that press is Dave. Thirty-one years. He can hear that bearing going before it goes. He's out in June of next year and I have nobody behind him. I can't hire a maintenance tech at the wage band corporate approved. That req's been open five months.
Rep
So the knowledge that's been protecting line 3 walks out in eighteen months. Has anybody above you connected those two things — the retirement and the unplanned hours?
Buyer
No. And that's the conversation I'd rather have before it happens than after. But I'll tell you what'll kill this — we did sensors in 2022. Little vibration pucks. Generated a pile of alerts into a shared maintenance inbox, nobody acted on any of them, it was dead in six weeks. My team's reactive all day. They don't have time for another tool.
Rep
The shared inbox is the failure, every time. An alert that isn't a work order in the CMMS with a named owner and a spare on the shelf is noise, and maintenance learns to ignore it inside three weeks. Let me ask you the thing I ask before I talk about any product: at 2am on second shift, if your on-call got a notice that said 'main drive bearing, nine to fourteen days,' what would he actually do with it?
Buyer
Honestly? He'd tell the supervisor and it'd get talked about in the morning meeting. And then it'd depend on whether the bearing's on the shelf, which — critical spares got squeezed last year for working capital.
Rep
Then that's the real project, and it's bigger than a sensor. Here's what I think is worth 45 minutes: you, your maintenance manager, and whoever owns IT/OT security. We pull twelve months of downtime by asset on line 3 and the two behind it, we settle the twenty-two hundred number with your controller's input, and we build the payback so it's ready when the AR window opens in Q3 instead of you writing it from memory in August. If it doesn't clear your hurdle rate I'll tell you on that call. Tuesday at 7 before your production meeting, or Thursday at 2?
Buyer
Tuesday at 7. Bring the format you want the downtime data in and I'll have my reliability guy pull it.
Objections you will hear
What they say, and what you say back.
| Objection | How to answer it |
|---|---|
| “Capex is locked for the year — come back and see me next budget cycle.” | Fair, and I'm not going to pretend I can unlock it. Two questions. First: is this shaped like capex, or can it run as a monthly opex line under MRO or controllable spend? That's usually a different approval path and often inside your signing authority. Second: if we're talking about next year's AR anyway, the work of building it — the downtime baseline on your worst three assets, the payback math your controller will actually sign — has to happen now, or you'll be writing it from memory in August. Let me help you build the number. If it doesn't clear your hurdle rate, you throw it out and I stop calling. |
| “The floor barely uses the systems we already have. Why would this be different?” | That's the most honest objection I get, and it usually means the last system asked operators to enter data. So ask me: who on your floor has to do something new for this to work? The answer should be nobody. If the value depends on an operator remembering to log a stop at shift change, throw us out. What I want to know from you is which system went unused and why — because if it was a change management failure rather than a tooling failure, we'll fail the same way and I'd rather find that out today than in month four. |
| “We already have a CMMS and a PM program. We're on schedule with our PMs.” | Good — most plants I walk aren't. But PM compliance and unplanned downtime aren't the same metric. If you're at 95 percent compliance and still eating unplanned hours on the bottleneck, that tells you the intervals are calendar-based rather than condition-based: you're changing bearings that had life left and missing the ones that didn't. What's your MTBF trend on your top three assets over the last year? If it's flat or getting worse while PM compliance is high, that gap is the whole conversation. |
| “You're not putting anything on my network or touching my PLCs. IT will never approve it.” | Understood, and I'd be worried if you said yes fast. This doesn't touch the control layer — no PLC connection, read-only, and it can run on a separate cellular gateway that never sees your plant network. What I'd ask for is fifteen minutes with your IT/OT security person early rather than late. If there's a hard blocker I want to hit it in week one, not after your reliability team has spent two months building a business case that dies at the security review. |
| “We tried a sensor pilot two years ago. It generated a bunch of alerts nobody acted on and it quietly died.” | What happened to the alerts — did they go to a shared inbox? That's almost always the failure. An alert that isn't a work order in the CMMS with a named owner and a spare on the shelf is just noise, and maintenance learns to ignore it inside three weeks. Before we talk product: tell me what your on-call would actually do at 2am on second shift with a notice that says this bearing fails in nine days. If there's no clean answer, we should fix that first, with or without me in the room. |
| “My maintenance team is already stretched. They're reactive all day — they don't have time for another tool.” | That's exactly the trap. Reactive teams stay reactive because every unplanned failure eats the hours you'd need to get ahead of the next one. I'm not asking for more of their time — I'm asking whether moving four of last month's emergency callouts into planned weekend work frees up enough wrench time to matter. What was your planned-versus-reactive work order ratio last quarter? |
| “Send me some information and I'll take a look.” | I'll send it, but a PDF won't give you anything you can put in an AR. What would be worth your time is fifteen minutes where you tell me your top three downtime assets and what a shutdown hour costs on each — you almost certainly have that number for the constraint. If the math doesn't get to a payback under eighteen months, I'll say so on that call and stop bothering you. |
Questions reps ask about this call
- What are the best discovery call questions for manufacturing prospects?
The ones that move from symptom to mechanism. Start with "which asset does everything route through, and what's it doing on a bad week?" then follow with "walk me through the last unplanned event on it" and "what did the downtime code say, and who decided that?" Most plants can produce unplanned downtime hours by asset instantly and cannot produce a cause — that gap is where the conversation gets real. Then quantify: unplanned hours per month on the constraint, cost per downtime hour, planned-versus-reactive work order ratio, overtime as a percent of direct labor, escapes and customer PPM last quarter.
- How do I ask about budget when they tell me capex is locked?
Don't ask about budget at all — ask about path. "What's your signing authority before this has to go to an AR?" and "where are you in the AR cycle right now?" tell you more than any budget question. Most plant managers have $25K or $50K of authority and everything above it goes into an appropriation request submitted in Q3 for the following year, competing against a new press and a roof. If capex is locked, your two real moves are testing whether it can run as opex under MRO or controllable spend, and offering to help build next year's payback number now instead of in August.
- How should the call differ between a Plant Manager, a Maintenance Manager and a VP of Manufacturing?
The Plant Manager owns OEE on the constraint, OTIF and the Saturday overtime, and he's the one on the phone with the customer's supplier quality engineer. The Maintenance Manager or Reliability Engineering Manager owns MTBF, MTTR, PM compliance and the planned-versus-reactive ratio — go straight at wrench time and callouts. The VP of Manufacturing or VP of Operations thinks across plants: cost per unit, standardization, and whether a fix at one site is repeatable at six. A Quality Manager's whole world is escapes, 8Ds, first pass yield and the threat of controlled shipping level 1. Same product, four completely different opening questions.
- They keep answering in one-word sentences. What am I doing wrong?
Usually one of three things. You pitched at layer one — they said "downtime is a problem" and you said "we solve that," which ends the thread. You stacked questions, so they answered the easiest one. Or you asked a leading confirmation question like "that's probably costing you a lot, right?", which earns a "yeah" and teaches you nothing they'd repeat to their controller. Fix: one open question, three follow-ups on the same thread before you change subject, and three full seconds of silence after anything heavy. The sentence about the millwright who retires in June comes out in that silence, not in your next question.
- What's a legitimate next step to close a manufacturing discovery call on?
Not "I'll send over some information." Prescribe a working session with named people and a named output: 45 minutes with the Plant Manager, the Maintenance Manager and whoever owns IT/OT security, to pull twelve months of downtime hours on the top three assets and build the payback number for the AR. Offer two specific times, get the invite out before you hang up, and confirm out loud what they're bringing — downtime by asset and their cost-per-downtime-hour. A next step with data homework attached survives; a demo invitation usually doesn't.
- How do I get better at this before a real call?
Rehearse against a difficult buyer, not a friendly one. The manufacturing discovery call fails in predictable places: the minute-six "so what do you guys actually do," the 95-percent-PM-compliance deflection, the dead 2022 sensor pilot, and the capex wall in the first ninety seconds. DrillCall lets you run those as voice roleplays with a Plant Manager or Maintenance Manager persona who pushes back the way a real one does, so the first time you hear "we tried sensors two years ago and it quietly died," it isn't on a live call with a $200K opportunity attached.