-
The quote that looked fine until month four
-
What you think the problem is
-
The deeper reason: you’re pricing a component, not a capability
-
The deeper reason people miss: sourcing gets treated as a purchasing task
-
What the gap actually costs
-
What I do now, before any sourcing quote gets signed
-
The short version
The quote that looked fine until month four
In Q3 2023 we replaced two aging lines. I had four quotes in a spreadsheet for what was, on paper, the same job: an SMD pick and place machine that could handle our board mix at around 20,000 components per hour. The spread was wider than I expected — $71,000 at the low end, $118,000 at the high end.
The low quote came from a sourcing agent. Clean PDF, three references, 30-day lead time. My controller liked it. I liked it.
What that $71,000 didn’t cover: feeder banks (quoted separately, “at market rate”), the nozzle set beyond a standard six (we needed eleven), the feeder cart, the offline programming seat, or the commissioning visit. By the time the line was actually running we were at $96,400. We also lost nine working days waiting on a nozzle that shipped from overseas.
I’ve been doing procurement for a mid-size contract electronics manufacturer — about 140 people between the floor and the office — for nine years. I’ve negotiated with more than 40 equipment vendors and logged every one of those purchases in our cost tracking system. And I still nearly made this mistake.
Most people researching SMT pick and place machine sourcing are asking the wrong first question. They’re asking who has the lowest price. The right first question is what does this price actually cover, and what comes back six weeks later as a separate invoice.
What you think the problem is
If you’re buying a PCB pick and place machine, the natural comparison is unit price. You take the specs — placement rate, feeder count, board size, component range — you line up the quotes, and you pick the one that gives you the most capability per dollar.
That works fine for commodity items. It works badly for placement equipment, because the machine on the pallet isn’t the thing you’re buying. You’re buying a capability that only exists once the machine is integrated with everything around it.
Here’s the trap. Two quotes for “the same” machine are almost never for the same machine. One includes a 12-zone feeder bank; the other assumes you’ll reuse feeders from the old line. One includes a two-day on-site commissioning; the other includes a PDF and a phone number. Both say pick and place machine at the top. The unit price comparison is real — it’s just comparing two different scopes.
The deeper reason: you’re pricing a component, not a capability
Placement rate is the number everyone compares. It’s also the number most likely to have been measured under conditions you’ll never reproduce.
IPC-9850 defines how surface mount placement equipment is characterized — and the key word there is characterized. A headline figure is only valid under the specific conditions stated alongside it. If a vendor quotes you 45,000 CPH without telling you the board size, the component mix, the feeder type, or the travel distance involved, you have a marketing number, not a spec. Ask for the test conditions in writing. (Source: IPC-9850, Surface Mount Placement Equipment Characterization; confirm current revision and scope before relying on any single figure.)
This matters for sourcing because it means you can’t rank pick and place machine manufacturers on one number. A machine that hits 45,000 CPH on an optimal board may hit 14,000 on yours if your mix is heavy on small passives with wide travel. The capability you’re actually buying is your throughput. Nobody quotes that unless you force them to.
The other half of the capability is what sits upstream and downstream. A placement machine is a node in a chain: loader, printer, SPI, placer, reflow, AOI, unloader. If your conveyor interface height is 920mm and the machine ships set for 880mm, that isn’t a spec mismatch. That’s a line-down situation until someone delivers a kit.
The deeper reason people miss: sourcing gets treated as a purchasing task
Here’s the part I got wrong for years.
When a line needs replacing, the pressure lands on the purchase order, not on the integration. The budget cycle closes. The production date is set. So sourcing gets compressed into get three quotes and pick one — which feels rigorous but isn’t, because all three quotes are answering a question you haven’t fully defined.
It took me about four years and roughly 30 equipment purchases to understand that the expensive part of a machine is rarely the machine. It’s the fit.
We didn’t have a formal pre-purchase verification process for a long time. It cost us on that $71,000 machine. It cost us again two years later when we bought a used placer that turned out to need a firmware unlock the seller didn’t own the rights to — a $4,200 surprise plus a two-week delay. The third time something like this happened, I finally built the checklist below. I should have built it after the first time.
What the gap actually costs
Let me put real structure on this, because “hidden costs” is a phrase that means nothing until you itemize it.
For a mid-speed SMD pick and place machine, here’s what sat outside the base price on our last three purchases:
- Feeders. Usually the largest line item after the machine itself. A working line needs 30–60 of them. Depending on type and condition, that’s a five-figure addition on its own.
- Nozzles. A standard set covers common sizes. Your component mix probably doesn’t. Individual nozzles are small; a full custom set is not.
- Feeder carts and trolleys. Often assumed on the buyer’s side, rarely included on the vendor’s.
- Software. Offline programming seats, license terms, and whether the license transfers if you resell the machine. Ask now, not in year three.
- Commissioning and training. A machine that arrives and isn’t commissioned is a very expensive piece of furniture.
- Spare parts. What’s the lead time on the three parts most likely to fail? If the answer is “three weeks from overseas,” that’s your real downtime exposure.
Add those up against a $71,000 quote and you land at $96,400. That’s a 36% gap between the number on the PO and the number that hits your cost center.
Now add the downtime. Nine days of a stopped line isn’t just nine days of no output. It’s the orders you pushed, the overtime you paid to catch up, and the customer conversation you had to have. I don’t have a universal formula for that number — it depends on your margin per board and your order book (mental note: someone should build that calculator). But you have the inputs. Run it before you sign, not after.
What I do now, before any sourcing quote gets signed
Short list. The list is short on purpose — the point is to actually run it.
- Ask every vendor to quote the delivered, commissioned, producing state. Not the machine. The state.
- Require an itemized exclusions list in writing. “What’s not included” tells you more than “what is.”
- Get the placement rate with its test conditions attached, per IPC-9850 or an equivalent characterization standard.
- Confirm feeder count, feeder type, and feeder condition — new, refurbished, or your existing stock.
- Send your actual component list and ask them to confirm nozzle coverage in writing.
- Ask for the three highest-failure spare parts, with current lead times and prices.
- Confirm conveyor interface height, footprint, power, and air requirements against your actual floor layout.
- Ask whether the software license is perpetual or subscription, and whether it transfers.
- Put a hold-back in the payment terms tied to the line producing your boards at your cycle time.
Nine questions. Maybe two hours of work, spread across a week.
The short version
I’m not a process engineer, so I can’t tell you whether a specific head configuration will hold your smallest components at your cycle time. That’s a question for your process team, and you should ask them early rather than after the PO goes out. What I can tell you from the procurement side is that the difference between a good machine purchase and a bad one is almost never the machine.
It’s whether you defined the scope before you compared the prices.
To be fair, running this checklist takes longer up front, and sometimes it means missing a price window that other buyers are chasing. I get why people skip it — budget deadlines are real. But I’ve now done this both ways, and the version where you spend a week asking questions is cheaper than the version where you spend two months fixing a line. Every time.
Prices and lead times above come from our own purchase records between Q3 2023 and Q4 2024. Equipment pricing varies heavily by configuration, condition, region, and market timing — treat any figure here as a reference point, not a quote, and verify current pricing directly with vendors before you budget.