Make room for public clouds in your industrial IoT strategy

Looking across this market, vendors of industrial IoT software platforms employ one of five broad strategies to the cloud.

The Rise of Industrial IoT

Not too long ago, executives from manufacturing and related industries would not consider running their internet of things (IoT) initiatives from a cloud. Now? It's hard to find one who doesn't enthusiastically embrace some role for the public cloud. Makers of industrial IoT (IIoT) software platforms have undergone a similar change of heart, pivoting from making grand plans for multibillion-dollar global data center rollouts to wholeheartedly partnering with one or more of the hyperscale public clouds.

Top Cloud Providers

Top cloud providers: AWS, Microsoft Azure, and Google Cloud, hybrid, SaaS players

Here's a look at how the cloud leaders stack up, the hybrid market, and the SaaS players that run your company as well as their latest strategic moves.

Read More

So far, so simple: Public cloud is now part of the industrial IoT story. But there are a multitude of ways to use a public cloud, from just installing your own software there to fully integrating with all the cloud, development, and IoT services that the hyperscale public cloud platforms increasingly offer. I explore this range of options in my latest report.

Looking across this market, we observe that vendors of industrial IoT software platforms (and their customers) employ one of five broad strategies to the cloud. They:

  • Achieve global reach by offering software for installation on cloud virtual machines. This strategy treats the public cloud as nothing more than a source of infrastructure services (compute, storage, networking) on which vendors simply install their existing software. It's an easy way to push software into new geographies where they may not have a local presence and an easy way to offer services on clouds with which they are not more closely aligned.
  • Avoid undifferentiated heavy lifting by using key services from their host cloud. With a little bit more effort, IIoT software platform vendors can tweak their software to use key services from each cloud on which they run. These services include things like load balancing, message queuing, and basic data management.
  • Accelerate innovation by integrating deeply with their host cloud. Every hyperscale cloud is different, and every hyperscale cloud has hundreds of services, with new ones added all the time. With this strategy, IIoT software platform vendors link themselves tightly to their cloud(s) of choice, benefiting from all of the cloud provider's own R&D spend and bringing rich customer-facing solutions to market more quickly. This can be a difficult strategy to maintain across more than one cloud, and it also makes the vendor reliant on their chosen cloud's future roadmap, pricing, and strategic direction.
  • Reduce duplication of effort by using pieces of their host cloud's IoT functionality. None of the hyperscale public cloud platforms only offer infrastructure services anymore. All of them are entering new markets and use cases, and all of them have some form of IoT-specific capability to offer. With this scenario, IIoT software platform vendors use some of those capabilities too, simplifying basic things like device management or routine data analysis.
  • Focus on unique capabilities by integrating with their host cloud's IoT functionality. Taking integration to its logical extreme, some vendors of IIoT software platforms partner strongly with a single public cloud provider. They use all of that cloud's infrastructure services, and they use all of its IoT and analytics capabilities. They focus on providing a thin — but valuable — layer of differentiation on top of that and work hard to make prospects and customers understand the additional value that they offer. It's a good strategy, and a sensible strategy, but also the one that's most at risk if the provider of the underlying cloud decides to focus elsewhere or — even more worryingly — decides to focus on adding exactly the features that were previously your layer of differentiation.

This post was written by Principal Analyst Paul Miller, and it originally appeared here.