I’ve Climbed Three Mountains Looking for the IoT Summit. Here’s What I Finally Figured Out.

Three decades across engineering services, component distribution, and cloud infrastructure taught me where the industry kept placing its bets and why the real answer was never at the top of any single mountain.

It’s been called M2M, IoT, the Internet of Everything, IIoT, and the Fourth Industrial Revolution. Hardware companies, software vendors, cloud providers, consultants, and systems integrators have all claimed to have the key that would unlock this multi-trillion-dollar market. Some of the largest, best-resourced companies in the world made serious runs at it. A lot of them have quietly retreated.

I’m not here to say they were wrong to try. I was right there with them, on more than one of those bets. Each one opened real markets, created real revenue, and delivered genuine value. But none of them — not a single one — delivered what this industry has been promising since the first M2M modem went into the field: real digital and business transformation driven by connected products and operations.

I’ve spent thirty years figuring out why. Here’s what I think I finally understand.

Mountain One: Engineering Design Services

In the early days of M2M, building a connected product was genuinely hard. Embedded systems, radio modules, proprietary protocols, custom gateways, none of it came off the shelf. Engineering design services firms became essential because the knowledge to assemble a connected product didn’t exist inside most product companies. We were the people you hired when you didn’t know how to build the thing.

The bet made sense: expertise was the scarce resource, and whoever owned it would be indispensable.

That was true until it wasn’t. Ecosystems matured. Tooling standardized. The hard knowledge diffused into the market, and what was once a moat became a commodity. Expertise alone was never going to be the long-term answer. But it gave the industry something important: the first generation of connected products, and something arguably more dangerous: proof that the thing could be done.

Mountain Two: Component Distribution

As the building blocks of connected products became more standardized, the bet shifted. If you could aggregate and distribute components at scale, the right chips, modules, antennas, sensors you’d own the supply chain. And if you owned the supply chain, maybe you’d eventually own the solution.

The world’s largest distributors came into M2M/IoT with real advantages: catalog depth, customer relationships, global logistics. They weren’t wrong about the opportunity. What they got wrong — what we all got wrong — was thinking that access to better parts solved the real problem.

The market didn’t need a faster path to a working chip. It needed a path to operating ten million devices after the chips were deployed. A component strategy gets you to the point of connection. What nobody wanted to say out loud was that's also where the customer's problem was just beginning.

Mountain Three: Cloud Infrastructure

Then the hyperscalers arrived. If the challenge of connected products was data at scale storage, compute, connectivity management well, that sounds like infrastructure. And infrastructure is something the cloud does better than anyone.

In some ways they were right. The cloud is essential. You cannot run a modern connected operation without it. But infrastructure doesn’t make decisions. Compute doesn’t figure out what should happen when a firmware update fails across 40,000 devices simultaneously. It doesn’t know that one of those devices is in a hospital, another is in a moving vehicle, and a third is under a service contract with a different SLA than the other two.

That kind of operational intelligence isn't a cloud feature. It never was. What this industry kept building were better mountains. What it needed was something that ran between them.

What All Three Mountains Have in Common

Here's the thing I had to sit with for a while: none of these were bad bets. Every layer genuinely matters. You can't build an intelligent connected operation without good hardware design. You can't scale without reliable supply. You can't process millions of data streams without cloud infrastructure.

But every one of these bets was a bet on a single layer of the stack. And the pattern held across all three. The mistake wasn't ignorance. Every one of these players understood their layer deeply. The mistake was believing that depth in one dimension could substitute for breadth across all of them. It doesn't work that way. It never did.

What this industry actually looks like, when you step back far enough to see it honestly, is a mountain range. Not a single summit. Engineering design services is a mountain. Distribution is a mountain. Cloud is a mountain. Enterprise applications your CRMs, ERPs, field service tools are mountains. None of them goes away.

The question that nobody answered well for thirty years is: what links them?

The Answer Isn’t Another Mountain

The answer isn't a better chip, a smarter distributor, or a more powerful cloud region. It's the discipline that sits above all of those layers and makes them work together.

Whether you're building connected products or connecting your operations, the stack decisions you face are the same and nobody has made them easy. What do you build bespoke? What do you buy off the shelf? How much hardware diversity is manageable? When do you use an integrator, and when does that become a trap paying someone to stitch together things that shouldn't require stitching?

The honest answer is that there is no universal hardware foundation. There is no single connectivity standard that wins everything. Enterprise applications don't become IoT platforms just because you run devices through them, but you'd be foolish to throw them away either. They're part of the value chain. The stack is always going to be heterogeneous. That's not a problem to solve. That's the reality to operate within.

What's been missing and what the industry now has is exactly that discipline. Not another point solution, not another consultant's integration project. An orchestration layer: workflow automation and policy execution that bridges legacy systems with new ones, connects device signals to business rules, and allows organizations to operate across a diverse and evolving stack without ripping everything out every three years.

Orchestration doesn't replace the mountains. It connects them. It lets you preserve the investments you've made, add new capabilities without starting over, and operate at a scale and consistency that no amount of human coordination or custom integration can match.

The reason I’m at EdgeIQ is simple: After thirty years of climbing different mountains, I know what it looks like when someone finally builds the path between them.

-

Joseph Zaloker

VP, Hyperscaler & AWS Partnerships · EdgeIQ

30+ years across M2M, IoT engineering, connected product distribution, and cloud infrastructure