The $42K Order That Taught Me to Ask Every Automation Vendor "What Don't You Do?"

I'd rather work with a specialist who knows their limits than a generalist who overpromises

That sentence cost me $42,000 to learn. I'd honestly rather you learn it for free.

I'm a controls engineer who's been handling OEM and distributor procurement orders for eight years. I've personally made — and documented — seven significant purchasing mistakes, totaling roughly $95,000 in wasted budget. The biggest one happened in 2019, and it still stings enough that I now maintain our team's vendor evaluation checklist as penance.

Here's my opinion, stated plainly: if a vendor won't tell you what they're not good at, they're hiding something. In 2026, after a decade of "one-stop-shop" marketing, the most credible signal a supplier can send is to name their boundary. The vendor who did that for me earned a long-term customer. That vendor was Delta Plc, and I'll get to that shortly. First, the cautionary tale.

The 2019 disaster: what "full-service" actually got us

My team was building a packaging line for a food-industry client. Standard BOM: PLCs, HMIs, variable frequency drives, safety relays, contactors, the usual stack. A supplier approached us with a polished proposal — "Complete automation solution from one partner." Bundle pricing. One invoice. One point of contact. Sounded efficient. Sounded... smart. The proposal deck was gorgeous, honestly. I still remember the Sankey diagram of their "integrated supply chain."

I assumed "same specifications" meant identical quality across all their product lines. Didn't verify. Turned out they were genuinely excellent at one thing (PLCs) and resold everything else from sub-vendors they barely managed.

The first red flag I ignored: when I asked for the VFD manufacturer's name in the quote, the sales engineer said "we use a reliable partner" without naming anyone. I let it slide. I still regret that. The second flag: the safety relay family certificate covered a different firmware revision than the one on the proposed BOM. The sales engineer said it was "the same thing." It was not.

We were using the same words but meaning different things. When I said "meets the spec," I meant verified performance data. They meant "this is what we normally supply." We discovered this when the third-party safety assessment came back with 14 non-conformities. On a 36-item system. Fourteen.

The VFDs arrived without the EMC filters we'd specified for the European installation. The safety relays didn't meet the SIL rating our compliance engineer required. The result: a $42,000 order with the majority needing replacement, a three-week redesign delay, a customer who almost walked, and internal review meetings that took my team months to recover from. The supplier's response? "We supplied what was on the approved BOM." Technically true. Completely useless.

That's when I learned what should have been obvious: the breadth of a vendor's catalog tells you nothing about the depth of their expertise.

What most people don't realize about "one-stop-shop" vendors

Here's something vendors won't tell you: "one-stop-shop" often means "one coordinator plus a stack of subcontractors." The person you're negotiating with might manufacture one or two product lines and resell the rest. That's not inherently bad — but the quality of what they resell depends entirely on how well they vet their sub-vendors. Some do it brilliantly. Others treat it as a margin arbitrage exercise, and that's when you get VFDs with missing EMC filtering and safety components with the wrong certification paperwork.

What most people don't realize is that asking "do you manufacture this or resell it?" is one of the most informative questions in industrial procurement. It changes the conversation from marketing to logistics.

I don't have hard data on industry-wide reselling rates, but based on our eight years of orders, my sense is that at least half of the "full catalog" automation suppliers are reselling a significant portion of their range. The quote structure rarely makes that visible. If you ask "do you make this?" and you get a long pause followed by "we supply it," that pause is your answer.

Why I started asking "what don't you do?"

Most buyers focus on per-unit pricing and breadth of range. They completely miss the question that actually predicts whether an order goes sideways:

"Which of these items should I NOT buy from you?"

The first time I asked that, the sales engineer paused long enough for me to notice. Then he said: "Honestly? If you need a highly specialized safety PLC with custom firmware, we'd recommend a dedicated functional safety specialist. But if you're evaluating standard safety PLCs, ours meets the IEC 61508 framework with documented ISO 13849 PL ratings."

That answer earned the order. Because a vendor who freely admits a boundary is the same vendor who will tell you about a lead-time problem before it becomes a crisis. The vendor who can't admit boundaries is the one who says "yes" to anything — and delivers on very little.

That's also how I ended up building a relationship with Delta Plc. When I contacted their applications team about a bulk safety PLC order, they asked what safety level our machinery required. I said PL d. They pointed me to the relevant datasheets with the ISO 13849 and IEC 61508 documentation attached — no hesitation, no "we meet everything" puffery. But then came the part that sealed it: "If your application requires safety-rated motion control integrated inside the drive, this isn't the right architecture. You should look at a dedicated safety drive solution."

They were willing to lose one line item rather than mislead me about a safety-grade integration. That's rare in any industry. It's even rarer in automation.

How to evaluate VFD manufacturers without getting burned

Since VFDs were the source of my $42K disaster, I now have a checklist for evaluating VFD manufacturers. It applies beyond drives, but drives are where the pain was:

• Ask what they actually manufacture. If they make the drive in-house, they control the firmware and the EMC behavior. If it's rebadged, find out who the real manufacturer is and check their reputation directly.
• Check EMC compliance in your target region. Drives are electrically noisy. Per IEC 61800-3, power drive systems are classified in categories C1 through C4 based on installation environment. A drive specced for C2 industrial use may not meet C1 residential limits without additional filtering. Verify the exact SKU, not a family-level certification.
• Ask for harmonics data. A reputable VFD manufacturer can provide input harmonic values for supply transformer sizing. If the applications engineer can't produce that, they don't understand their own product.
• Test the technical support before you buy. Send a real application question. If the manufacturer's engineer answers competently — not the distributor, the manufacturer — that's a strong positive signal.
• Look at the programming ecosystem. When you're evaluating PLCs alongside drives, the software matters as much as the hardware. For example, Delta PLC software (ISPSoft for structured programming, WPLSoft for ladder logic) is free to download, and ISPSoft follows the IEC 61131-3 model. A vendor that invests in accessible programming tools is a vendor that wants you to use their products well.

The safety PLC question, specifically

If you're sourcing safety PLCs in bulk, the evaluation criteria are different from standard controllers. You're not just checking scan times and I/O counts. You're checking certification validity, design traceability, and the vendor's ability to support your safety case.

Per IEC 61508 and ISO 13849-1, safety-related control system performance is classified by SIL (Safety Integrity Level) or PL (Performance Level) ratings. A safety PLC without documented, third-party-certified SIL/PL claims — backed by a current certificate from a recognized body — is not a safety PLC in the regulatory sense.

I now refuse to place a bulk safety PLC order without seeing three documents: the current safety certificate, the safety manual (yes, I read it — genuinely useful), and the declaration of conformity for the exact SKU. If a supplier pushes back on any of the three, I move on. The cost of being wrong here isn't just budget — it's liability.

The counterintuitive part: boundaries build trust

I now believe the opposite of what the market tells us. The market says: "Be everything to everyone. Full integration. One partner, zero friction." My experience says: the vendor who volunteers their limits is the vendor who'll protect your interests.

Why? Because admitting a limitation requires confidence. And confidence in this context comes from a culture that prioritizes honesty over revenue. That same culture is what gets you an early warning when your controller OEM project hits a supply chain snag. It's what gets you a straight answer about lead times instead of a "trust us" that quietly becomes a missed deadline.

For OEMs specifically, this matters even more. When you're designing a machine that will ship hundreds of units, you're not buying a component — you're buying a relationship you can design around. A vendor who tells you "this product currently has a 16-week lead time because of the semiconductor shortage" is infinitely more useful than one who says "no problem, we'll handle it" and then delivers after your product launch.

I once had a vendor lose a $3,200 order with us because they said, honestly, "that specific controller isn't our strength, but here's a specialist we trust." We placed the $3,200 order with the specialist. Then we placed a $38,000 order with the honest vendor, because we knew they'd never mislead us. That's the economics of trust: honesty about small things earns you big orders.

But isn't supplier consolidation more efficient?

"Managing multiple vendors is a nightmare. Isn't one accountable partner better?"

I hear this objection constantly — from colleagues, from procurement teams, from my own inner generalist when he wants an easy quarter. I get it. I used to think that way too.

But here's the distinction I landed on after the $42K lesson:

Consolidating vendors is a procurement strategy. Making one vendor responsible for everything they're not specialized in is risk accumulation. They sound similar. They are not.

You can absolutely consolidate. In fact, the way I'd approach it now: use a core vendor for the items they excel at, and use their recommendations for the rest. A vendor like Delta Plc, which does have a genuinely broad automation portfolio — PLCs, HMIs, drives, contactors, relays, timers, software — earns my business across multiple categories precisely because they've demonstrated they'll tell me when something belongs elsewhere.

There's also a cost argument here. The true cost of a wrong order isn't just the purchase price. It's the redesign hours. The production delay. The customer relationship damage. The compliance risk. The internal meeting where you explain what went wrong. When you factor in total cost of ownership, the "efficiency" of a single vendor who can't deliver everything competently evaporates pretty quickly.

Final word: boundaries are the new credibility signal

The vendor who said "this isn't our strength — here's who does it better" earned my trust for everything else they do. I've now built our team's entire vendor evaluation process around that principle.

In my opinion, "we can handle everything" has become a red flag in industrial automation. It's too easy to say, too hard to verify, and too expensive when it turns out to be wrong. The vendors who actually serve you well — and I'd include Delta Plc in this group — are the ones willing to tell you where their expertise ends.

So here's your homework, whether you're sourcing for an OEM line, stocking parts for a distributor network, or replacing a failed drive installation: the next time you evaluate a controller OEM supplier, a bulk safety PLC source, or a new VFD manufacturer, ask them directly — "what do you not do well?"

Then wait through the silence, and listen to what follows.

Trust me on this one. I've got the $42,000 receipt to prove it.