I'm going to say something that might sound strange coming from someone whose job is literally to buy things: I don't want a vendor who claims they can do everything. I've been managing procurement for a mid-sized industrial equipment manufacturer since 2020—processing roughly 60-80 orders a year across 8 different vendors for everything from LoRa modules to circuit protection components. And after five years and probably 400+ orders, here's what I've learned to value most: honesty about what you don't do well.
The Pitch That Makes Me Suspicious
When a sales rep from a new supplier starts their pitch with 'We're your one-stop shop for all IoT connectivity needs,' my internal alarm goes off. It's not that I don't appreciate convenience—I do. But in my experience, 'one-stop shop' usually translates to 'we do a few things well and the rest is subcontracted to someone we barely vet.'
Take Semtech, for example. They're obviously the go-to for LoRa technology—their LoRa transceivers like the SX1272 and SX1262 are industry standards for long-range low-power IoT. But I wouldn't ask them to help me with a cellular IoT gateway design that's better served by a module from a different specialist. That's not a knock on them; it's just recognizing that different problems need different tools. The surprise wasn't that Semtech couldn't be everything to everyone—it was how refreshing it was when a sales engineer straight up told me, 'For that use case, you're actually better off looking at [competitor].'
That conversation earned my trust for every other product I buy from them. Because now I know: when they say they're good at something, they mean it.
Three Hard Lessons That Changed How I Buy
It took me three years and about 150 orders to understand that vendor relationships matter more than vendor capabilities. But the lesson really hit home through three specific experiences:
1. The 'All-in-One' That Did Nothing Well (note to self: vet the full portfolio, not just the flagship)
In 2022, I consolidated our ordering for wireless modules and routers under one vendor who promised 'seamless integration across all platforms.' The LoRa modules were fine—they were rebranded Semtech chips, basically—but the routers were garbage. They had compatibility issues with our gateways, the management interface was buggy, and their support team kept blaming the gateway manufacturer. After six months of headaches and a $15,000 loss in productivity (I tracked it), I split the order back to two specialists. The modular specialist cost 8% more but their tech support actually understood the product. I'll never make that mistake again.
2. The Vendor Who Told Me to Go Elsewhere (I only believed this was a good sign after the second time it happened)
In 2023, I was sourcing ESD protection diodes for a new board design. A vendor I'd worked with before for signal integrity products (they're great with Gennum stuff) said, 'Actually, for this specific application and current rating, you should look at a different series from Semtech's RClamp line—here's the contact for their applications team.' They didn't even try to upsell me on something they didn't have. That simple act of redirection saved me probably 10 hours of research and a potential design revision. They now get all my protection component orders, even when I could technically get them cheaper elsewhere.
3. The Assumption That Cost Me $1,200 (reverse validation in action)
Everyone told me to always verify a vendor's specialization before committing to a large order. I only believed it after skipping that step once. I assumed a supplier who did excellent LoRa gateways (based on the SX1301 chip) could also handle the antenna design for a custom deployment. Spoiler: they couldn't. The antennas they provided missed the target frequency band by 4%, which I didn't catch until prototype testing. I ate the $1,200 cost of redesign out of my own departmental budget because my manager rightfully pointed out I hadn't done my due diligence when selecting the vendor. Costly lesson, but I haven't repeated it.
The Argument Against Specialization (and Why It Falls Flat)
I know the counter-argument: dealing with multiple vendors creates overhead. More invoices, different support contacts, incompatible processes. And I get it—I'm the one who reconciles those invoices. But here's the thing: the overhead of managing 8 vendors is less than the headache of dealing with a single vendor who's mediocre at 6 of the 8 things they do.
The real cost isn't the time spent managing vendor relationships. It's the time spent fixing mistakes from vendors who overpromised. According to a study cited by the FTC (ftc.gov), inaccurate advertising and unsubstantiated performance claims—like saying 'our module works with all gateways' when it doesn't—cost B2B buyers an average of 15% in hidden integration costs. My experience actually matches that number pretty closely.
How I Evaluate Vendors Now
After five years of managing these relationships, I've come to believe that the 'best' vendor is highly context-dependent. Here's my current framework, for what it's worth (and my experience is based on mid-range technical components for industrial IoT—your mileage may vary if you're buying something totally different):
Phase 1: Ask the uncomfortable question. In the first meeting, I ask: 'What type of project would you tell a client to go elsewhere for?' The vendors who give a thoughtful, honest answer get a second meeting. The ones who say 'we handle everything' get a mental asterisk.
Phase 2: Check specialization. I don't just look at their product list. I look at their application notes, their design resources, their engineering team's depth. Semtech's LoRa documentation is a good example—hundreds of pages of RF design guidance, reference designs, and calculator tools. That depth tells me they know their stuff. A vendor whose 'expertise' consists of repackaging five different OEM products with their logo? Less confidence.
Phase 3: Evaluate the 'no.' The best indicator of a vendor's reliability isn't their 'yes'—it's their 'no.' A thoughtful redirection to a competitor or an honest admission that their product isn't the right fit shows they value the relationship more than the transaction. I've had vendors lose a $5,000 order by being honest. But they've each earned tens of thousands in subsequent orders because I trust them.
One More Thing: On Compatibility Claims
I have a particular pet peeve with compatibility claims. When a vendor says 'works with all LoRaWAN gateways,' that's a red flag. LoRa is a robust technology, but real-world deployment depends on regional frequency plans, gateway configurations, and network server compatibility. I've never fully understood why some vendors make these blanket claims—my best guess is they're trying to reduce purchase hesitation. But for someone like me who's been burned by integration surprises, it has the opposite effect.
Honestly, I'm not 100% sure what the right approach is for vendors who want to simplify their messaging without oversimplifying. But I know what works for me as a buyer: specificity. 'Compatible with AWS IoT Core for LoRaWAN and ChirpStack v4+' is better than 'works with major platforms.'
The Bottom Line
I don't need a vendor who can do everything. I need a vendor who's exceptional at what they do and honest about what they don't do. The vendor who said 'that's not our strength—here's who does it better' earned my trust for everything else they sell. That trust, by the way, spans multiple departments in my company now—I've routed them for signal integrity components, circuit protection, and wireless modules across different product lines.
If you're a vendor reading this: don't be afraid to say 'I don't know' or 'that's not for us.' Be afraid to say 'yes' to something you can't deliver on. And if you're a buyer like me: pay less attention to the product catalog and more attention to how the vendor handles the conversation about what they can't do. That's where the real signal is.
Based on my experience managing procurement for a 200-person manufacturing firm across 3 locations. Take this with a grain of salt: my vendor count and order volume might differ significantly from larger or smaller operations.
