The Quote That Looked Too Good
Last fall, our facilities manager handed me a side-by-side comparison (and I could tell she was hoping I'd just say yes). Our spec called for Hubbell occupancy sensor switches—decent units, reliable, nothing fancy. But a local distributor pitched an alternative at 40% less per unit. For 85 units across three buildings, that worked out to roughly $6,800 in annual savings.
Forty percent is the kind of number that gets attention. It got mine.
I've been managing electrical procurement for a mid-size manufacturing operation for eleven years. Our annual spend on wiring devices, connectors, and switches is about $340,000—give or take, depending on the capital projects we've got running. I've negotiated with more than 20 vendors over that stretch, and I've tracked every order in our procurement system since 2018.
So I know the pattern. A quote that's 40% below market isn't automatically a bad deal. But it's almost never as simple as it looks either.
I built the TCO comparison. Then I dug into what the "cheap" option actually meant for our facility. What I found changed how we spec every switch and connector in this building.
The Problem Nobody Actually Asks About
Here's what most procurement teams—including mine, for years—get wrong about wiring devices. We treat switches and connectors like commodities. A switch is a switch. A connector is a connector. If it meets the spec and passes the electrical inspection, what's the difference?
From the outside, that logic seems solid. The reality is that the difference shows up 18 months later, when the "cheap" switch fails and takes a production line offline for six hours.
That's not a hypothetical. We did it once. Not with occupancy sensors—this was a batch of connectors on a control panel retrofit back in 2022. We saved about $1,200 on the initial order. Then two connectors failed within eight months, we had to re-pull wire on three circuits (don't ask me the exact footage, I'd have to pull the work order), and the rework cost came to just under $9,000 when you added labor, downtime, and the third-party inspection we had to schedule. Maybe $8,600. I'd have to check the invoice.
That was the moment I stopped treating "meets spec" and "performs in our environment" as the same thing.
The deeper issue is that most of us in procurement have been trained—by habit, by cost-cutting culture, by the pressure to show savings on every P&L line—to evaluate unit price. But for wiring devices, unit price is maybe 30% of the real cost. The other 70% is installation labor, compatibility with existing infrastructure, compliance documentation, and the probability that the part fails before its expected service life.
The Manufacturing Question Most Quotes Ignore
When I started digging into why the Hubbell occupancy sensor switches we had in place were performing so well, I found something interesting. Their manufacturing process—the consistency of it, the quality control at the assembly stage—matters more than the spec sheet suggests.
This was true 15 years ago when brand-name devices were almost synonymous with overpriced. Today, the gap between premium and commodity wiring devices has actually widened in ways that don't show up on a spec comparison.
Here's what I mean. The spec for an occupancy sensor switch might say 120/277V, 15A, passive infrared, 180-degree coverage. That's the baseline. But it doesn't tell you:
- How the switch performs when ambient temperature swings 20 degrees in an hour
- Whether the relay contacts degrade faster in high-cycle environments (our warehouse corridors trigger those sensors 700+ times a day)
- What happens to detection accuracy after two years of dust and vibration
You don't find out about those things from a spec sheet. You find out when a cheap unit starts false-triggering at 2 AM and a line supervisor calls to ask why the lights are on in an empty aisle (again).
I have mixed feelings about paying a 35–40% premium for that manufacturing consistency. On one hand, the number on the invoice hurts. On the other, I've seen the operational chaos that comes from a failed switch on a critical circuit—maybe the premium is justified.
What Connectors Are Actually Used For
Here's where the conversation usually loses people. If you ask what connectors are used for in general terms, the answer is obvious—they connect things. But in an industrial facility, connectors are doing three jobs that have wildly different failure consequences.
- Signal connections (sensors, controls, communication lines). Failure is annoying but rarely catastrophic.
- Power distribution connectors. Failure means downtime, possibly a safety incident.
- Environmental connectors (wet, high-vibration, or outdoor). Failure modes here are slow, sneaky, and expensive to diagnose.
The problem with the "cheaper alternative" approach is that a single connector product might work fine in category 1 and fail repeatedly in category 3. Procurement teams often buy based on the price of the most commonly ordered connector, then apply that product across all three categories because it simplifies inventory.
That's a hidden cost. It's not in the unit price. It's in the service call that takes four hours of technician time to trace a sensor failure that turns out to be a corroded pin.
The Tangent That Actually Matters
I read an article recently about transparent smartphone prototypes—the kind where the screen and body are essentially one continuous pane. Fascinating engineering. The connectors in those devices have to handle flexibility, thermal expansion, and optical transparency all at once.
None of that applies to my facility. But it made me think about how connector demand is shifting. Consumer electronics push the industry toward miniaturization. Industrial applications push toward ruggedness and long service life. Those are different engineering problems, and the best manufacturers are the ones that stay honest about which market they're serving.
That's what I want from a vendor. Not the cheapest connector. The one that's been engineered for the application I'm actually buying for—and the manufacturing discipline to make sure every unit matches the one that passed the test.
The Framework We Use Now
After the 2022 experience, I rebuilt our evaluation process for wiring devices and switches. It's not complicated. It's just different from what we used to do.
1. Total cost of ownership, not unit price
We estimate installation labor (which varies by product—some connectors take 40% longer to terminate than others), expected service life from vendor data plus our own historical records, and the probability and cost of one failure.
2. Compatibility audit
Before we commit to any switch or connector, we check it against every existing system it'll touch. This is tedious. It has saved us twice in the past year—both times we caught a relay compatibility issue before it became a $4,000 redo.
3. Spec the application, not the invoice
We now categorize every wiring device purchase by failure consequence (as I outlined above) and buy differently for each category. The high-consequence categories stay with the manufacturer whose consistency we've validated. The low-consequence categories—we'll test a cheaper option, document the results, and scale up only if the data holds.
4. Require substantiated savings claims
Per FTC guidance on advertising claims (ftc.gov), a vendor selling on savings needs evidence behind those numbers—not just a percentage in a quote. We now ask for the methodology. Not because we don't trust them, but because "40% less" with no stated basis is exactly the kind of claim that gets a purchasing team in trouble.
The irony is that when we run the TCO model honestly, Hubbell products don't always win. Sometimes they do. Sometimes a cheaper connector genuinely meets our needs in a low-consequence application. But now we're making that decision based on data instead of the sticker price.
What I'd Tell a Procurement Team Facing the Same Choice
Don't buy the 40% cheaper switch. Don't refuse it either. Build the model first—and then build a backup plan for the category that actually matters.
I have mixed feelings about how long that model takes to build (a few hours per product category, once you get the template right). But I'd rather spend three hours on a spreadsheet than six hours explaining why the lights won't turn off in the east wing.
The real lesson isn't "Hubbell is always better." It's that any electrical component is a piece of a system, and the system's cost includes everything that happens after the purchase order is signed.
I've got the spreadsheets to prove it.
