I'll say the thing that usually gets me uninvited from vendor lunches: the cheapest control module is the most expensive one you'll ever buy. Not always. But often enough that I changed my entire procurement process because of it.
I'm a procurement manager at a 37-person industrial services company. I've managed our automation and MRO parts budget, roughly $180,000 a year, for the past six years. I know how to stretch a dollar. I also know when stretching it costs you two more later.
If you are sourcing a digital output module, a transmission control unit, a safety I/O module, an ABB AI820, a Bently Nevada 3300 XL component, or an ethernet analog output module—for generator controls, turbines, or any other industrial panel—this article is for you.
The Invoice Is a Cover Charge
When I audited our 2023 spending, I pulled every module purchase and sorted it by two columns: 'initial price' and 'total cost to our shop.' It took a weekend. Actually, it took two weekends because my spreadsheet was a mess.
Nine of the 23 modules I bought that year came from 'great deal' sources—surplus dealers, liquidation listings, online marketplaces. Six of those nine needed a second purchase within 12 months. Some arrived dead. Some worked for a month. Some didn't communicate with the PLC because the firmware revision was wrong. The initial savings evaporated. On average, the 'deal' ended up costing 34% more than the new part would have, once shipping, troubleshooting, and re-purchase were counted.
I'm not a statistician. I'm not an engineer. I'm just the person who signed the purchase orders and then had to explain the budget overrun.
The invoice is not the cost. That's the entire article, if you want the short version.
Let Me Show You the Math
Here's a real example from a generator control panel service project last year.
The distributor quoted $350 for a digital output module, new, with technical support included. A surplus dealer quoted $180 for the same part number, listed as 'new surplus.' I'm not a brand snob. I wanted the $180 one to work. It didn't.
The module looked right. Photos matched. Wait—actually, the module itself was fine; the stored configuration was the problem. But the first sign was the same: no communication. We spent three hours chasing it. Three hours of a field tech's time at $90 an hour. Then we had to overnight the correct module for $95 in freight.
Total with the surplus dealer: $180 + $270 labor + $95 freight = $545. The distributor would have been $350. The most frustrating part wasn't the $545. It was the three hours I spent explaining to a project manager why I'd tried to save $170.
Do the math. I did—on a napkin, then in our cost tracking system. That's a 56% difference hidden in fine print. Actually, 55.7% if you're a nerd. The cheap module was not cheaper.
A Part Number Is Not a Specification
Here's the part that took me years to learn. And I still don't understand all of it. I'm not an electrical engineer, so I can't speak to signal conditioning, isolation, or why an ABB AI820 behaves differently in one rack versus another. What I can tell you, from a procurement perspective, is that part numbers lie.
A part number doesn't tell you the firmware revision. It doesn't tell you whether the module was pulled from a decommissioned line after a saltwater leak. It doesn't tell you whether the safety I/O module has the safety rating your application requires. It doesn't tell you if that Bently Nevada 3300 XL card was calibrated before it left the previous site.
The same logic applies to transmission control units and ethernet analog output modules. Network settings, protocol versions, and configuration data can be left over from the previous owner. If the seller doesn't know what those settings are, they're not selling a module. They're selling a mystery.
A listing can say 'OEM compatible.' What does that mean? Per FTC advertising guidance (ftc.gov), claims need evidence. My policy now: if a seller cannot show me a batch photo, a pull history, or a test report, 'compatible' is just a word. A part number is not a guarantee. What I mean is: it's a starting point, not a specification.
Before You Call Me a New-Parts Snob
I'm not saying surplus parts are evil. Some of my best purchases were surplus. Last year, we found an ethernet analog output module for a legacy line that the OEM no longer supports. It arrived, it worked, and it cost us a third of the obsolete-market price. Not ideal, but workable. In 2025, you take that win.
The difference is how confident the seller is about what they have. The sellers I trust do three things:
- They show photos of the actual unit, not a stock image.
- They tell you why it's available—decommissioned, removed from a running plant, or cleared out of a warehouse.
- They put a written return or warranty statement in writing, not just a promise in a message.
If a dealer can't do those three, I walk. The price stops mattering. It's not a bargain if you can't send it back.
'But What If There's No New Part Left?'
Fair objection. I've been there. The ABB AI820 isn't in some OEM catalogs anymore. Bently Nevada 3300 XL is still in service in a lot of plants, but new replacements are getting harder to source. You can't always buy new from the manufacturer.
In that case, change the risk equation. Don't just buy the cheapest listing. Buy from a supplier who can tell you the exact revision, offers a return period, and answers emails within a day. Test the unit on the bench before it goes into production. Keep a spare if the budget allows. If it's a safety I/O module for a process you can't afford to trip, get the manufacturer's modern equivalent or a qualified third-party refurbisher. Don't rely on the word of a stranger with a stock photo.
On the other hand, if this is for a test bench or a low-criticality application where downtime is a nuisance rather than a crisis, buy the cheap one. I would. Just don't call it a cost-saving decision. Call it what it is: a calculated gamble.
What I Actually Do Now
After tracking six years of invoices, I came up with a rule that sounds obvious but took a spreadsheet to convince me: price is the question you ask last, not first. First, ask what the part is, where it came from, what's verified, and who answers the phone when it fails. Then ask the price.
I still buy from surplus sellers. I still recommend them for the right situations. But I stopped pretending the lowest number on the screen is the lowest cost. It usually isn't. Sometimes it is. But the hidden costs—rework, downtime, expedited freight, engineers who don't have time for this—turn the 'deal' into a donation.
I built a simple cost calculator after the generator control panel fiasco. A spreadsheet, not fancy. It forces me to enter freight, labor, downtime, and re-purchase estimates before I approve any module purchase above $300. That one habit probably saved us more than any vendor discount.
That's my honest opinion. You don't have to agree. But if you're sourcing a digital output module, a transmission control unit, a safety I/O module, an ABB AI820, a Bently Nevada 3300 XL, or an ethernet analog output module, save yourself the lesson I paid for.
Check the spec. Check the seller. Check the return policy. Then pay the price that buys you the least amount of headache.
Leave a Reply