Five key take-aways from a 7-week IoT odyssey

I have spent the vast majority of the last seven weeks travelling for business, meeting clients and speaking at conferences. The itinerary has taken in Stockholm, Berlin, Amsterdam, Las Vegas, San Francisco and Seattle (and environs and a few other places in between). The beauty of spending so long with clients and at events is that it allows me to take the pulse of the industry, to understand what’s happening. 

In this blog post I set out five areas in which talking to dozens of organisations deeply involved in the Internet of Things has shaped my thinking about how the IoT space is evolving. There are many more trends, including cloud integration, edge compute, private networks, LEO satellites and much more besides. But here are five that stood out for me, all of which are quite closely interlinked.

The need for ‘credible IoT’

Not all IoT vendors are equal. Whatever the aspect of IoT you look at there tends to be a plethora of different vendors, which on paper seem to have functionality which is comparable and fitting into one of the numerous well-established categories of IoT provider (more on this later!). However, the reality is that there is typically a world of difference between them. It may even be the case that many of them ‘can’ replicate what their competitors do, but don’t have the robust, industrial-grade, tried-and-tested solutions, or years of hard-earned experience, of some of their peers. This type of dynamic is particularly true in the provision of connectivity, IoT platforms, and systems integration/consulting.  

To give one example, back in November last year we tracked 175 companies (and there are more) which fit under the broad umbrella of ‘IoT MVNO’. At the recent conferences, particularly at Mobile World Congress Las Vegas, and in other discussions, I have spoken a lot with different MVNOs about their capabilities. One thing that emerges is that there is a set of operators that have earned their stripes, providing connectivity to demanding customers in complex environments. Take, for instance, OptConnect, which has a heritage in connecting ATMs, Kajeet, which built capabilities in policy management in connecting devices in the education and healthcare sector, or Datablaze, which has 20 years’ experience providing customised hardware with connectivity. And contrast these players with newly arrived IoT MVNOs that are doing little but reselling connectivity based on a skim on an MVNE (Mobile Virtual Network Enabler) platform. These things are not the same. 

I’m reminded of Plat.One, formerly Abo Data, which was acquired by SAP in 2016. There was a company that had built, over decades, expertise in industrial control systems, which fitted neatly into the IoT Application Enablement Platform (AEP) space that emerged in the 2010s. That tried-and-tested industrial-grade solution sat alongside many upstart AEPs built in just a few months specifically as ‘IoT’ platforms. 

Why does this matter? Quite simply because enterprises procuring capabilities will have limited knowledge and understanding about what marks out an experienced vendor from a fake-it-til-you-make-it competitor. What’s needed is ‘credible IoT’. Our advice is always this: look for, first and foremost, a heritage in the space. Recently an enterprise engaged us because the team were unhappy with their IoT connectivity provider and wanted recommendations on alternatives. The underlying reason they were unhappy was that the provider they had chosen had an inappropriate solution for their needs, although on paper would have qualified as a vendor of the type of IoT connectivity that they were looking for. They even asked us if they had chosen a good vendor. The answer to which was a resounding “yes”. But not for the type of deployment they needed. 

The lesson is simple: there is no single ‘best’ global IoT connectivity provider applicable to every scenario. And even the good ones might not be appropriate. Would I pick Singtel, whom I also sat down with at the IoT Tech Expo in Amsterdam, as my provider of regional connectivity in Asia? Absolutely. Would I pick them for a multi-regional deal across Europe and North America? No. And they probably wouldn’t want that deal anyway. In our forthcoming CSP IoT Peer Benchmarking Report we will reflect that requirement that the selection of the ‘best’ provider is typically a regional thing. 

Another company that I sat down with was Twilio to discuss its relatively new Super SIM product. The interesting thing about Twilio is that it has a very strong heritage in what I might term ‘the action layer’, i.e. once you’ve received the trigger, what do you do? That’s its bread and butter in the A2P messaging space. In IoT it’s about triggering a trouble ticket to be raised, or a message sent to a field service technician, or whatever. Twilio has great heritage there and if it can transition it to the IoT space it’ll do quite well. Credibility comes in many forms.

Essentially it boils down to this: as an enterprise, your needs are specific, make sure the vendor can credibly claim to address them. If in doubt, ask us!

A need for ‘managed IoT’ and cross-optimisation

An adjunct to the need for credible IoT is a requirement for ‘managed IoT’. The idea that you can just throw a set of hardware, connectivity, platforms etc. at a customer and they come out with a winning IoT capability is pie in the sky. Every organisation selling into IoT needs, in some way, to provide a degree of consultancy, or enriched post-sales service, call it what you will. 

I’ve spoken in articles and on the Wireless Noodle podcast before about the need also for cross-optimisation, which is part of this managed approach. With IoT being deployed in increasingly constrained environments, particularly courtesy of the LPWA technologies, as well as other hardware, protocols and software optimised for IoT (the so-called ‘Thin IoT’ stack there is an increasing requirement for the various constituent parts of IoT solutions to be optimised with each other. Not just bundled (as I had the conversation with the CEO of one IoT hardware manufacturer) but cross-optimised, ensuring the right network, service providers, protocols etc. are being used and that the application is being architected as appropriate. 

One of the interesting current trends, which comes out in the discussions I’ve been having, is that there is a tremendous asset for service providers in having a consulting/systems integration element to what they do. The likes of NTT and Orange are treating IoT as a functional element upon which to layer on a deeper consulting engagement. They are very definitely not just selling connectivity.

Most enterprises need a managed solution, not just a set of building blocks, and part of that management is in ensuring that the various constituent parts work together. As it happens, the organisations best able to deliver these managed capabilities also tend to be the ones that fit the criteria of ‘credible’ in the previous point too. Not surprising.

I thought long and hard about the right term for this, and ‘managed’ has a very 1990s feel to it, when managed services were all the rage. Perhaps I’ll think of a better term some time, but for now managed seems to be about right.

From IoT value chain to a new IoT taxonomy

Another thing that is becoming increasingly apparent from discussions with the great and the good in the IoT space is that it is evident that we need a new taxonomy for the IoT space, exploding the old ‘IoT value chain’.

Historically, there has been a lot of discussion about the ‘IoT value chain’. The reality is that it’s not a value chain at all. In a value chain A sells to B, which adds value and then sells to C, and so on. In IoT there are an assortment of components, which are built together to form a product or solution. But the entry point to the value chain could be anywhere. For instance, you could procure your hardware first and then add connectivity, or you could go to a cloud provider or systems integrator or full solution provider and start your IoT journey there. None of this looks like a value chain at all.

Notwithstanding the fact that the old terminology is wrong, more importantly, the old taxonomy of organisations is also increasingly wrong. The supposed value chain comprised semiconductors, modules, devices, network operators, perhaps an MVNO, connectivity management platform, application enablement platform, integration/consulting and the end solution. More or less each of those elements had a set of vendors associated with them. There were a few spanning multiple elements but it was the exception. However, today, those roles are less and less an accurate reflection of what organisations actually do. 

In the hardware space we see more and more silicon vendors moving further into the hardware stack. One outlier example of this is the Semtech move to acquire Sierra Wireless. We also see the big IoT hardware companies moving more strongly into connectivity (e.g. Telit and Quectel), plus companies like Blues Wireless blurring the lines between hardware and connectivity. This will be further exacerbated by eSIM and then iSIM, whereby the connectivity is baked into the device. 

Elsewhere, we often hear connectivity vendors wanting to position themselves more as connectivity platform providers. For instance, the likes of Eseye, an established MVNO, allowing customers to bring their own connectivity. Or floLIVE which straddles the world of connectivity management platform and IoT MVNO (plus private network provider) in a different way. We also see other connectivity providers, such as Vodafone and Telia much more positioning themselves around managing outcomes, for instance managing operational assets or offering insights. 

The increasing importance of the cloud hyperscalers will also play a role in changing this taxonomy. Functionality that was previously discrete, such as mobile core networks, has become software functions on cloud platforms. Plus, those cloud hyperscalers, either organically or through acquisition, are spanning many of the functions across IoT themselves. Whatever happens, whether they are acting as enablers of scale for other vendors or taking them on head-on, the hyperscalers will change IoT. Edge computing is also playing a significant role in upsetting the existing architectures, changing the vendor environment as a result. 

What is needed is a new taxonomy of IoT vendors, better reflecting these and other changes in the space. We’re working on it.

The IoT app store?

Over the years there have been quite a few attempts at building an ‘IoT marketplace’, often using the analogy of an app store. Typically, it has been Communications Service Providers attempting to do so, but in some cases other vendors. The problem with all of these was scale. Is it worth an application developer, or even a hardware or services provider, going to the effort of integrating with one marketplace, with all the certification and approvals entailed? Perhaps, perhaps not. 

What this space has been crying out for is rationalising to a handful of global scale marketplaces. During his keynote presentation at IoT Tech Expo North America, Amir Kashani of Stanley Black & Decker bemoaned the fact that there aren’t such marketplaces where his company can find and test multiple vendors in a particular area. Everything is highly fragmented and manual to pull together. 

What has been interesting, therefore, has been a series of conversations with some of the bigger global IoT players, and I’m thinking particularly about AWS, Microsoft, Oracle and Siemens, about creating just this kind of ecosystem, or something similar. These are the biggest companies in Digital Transformation products (whereas other major vendors such as Accenture and IBM tend to focus more on services). If anyone is able to catalyse such an initiative, it is them. 

And, indeed, they have been. Siemens launched (or, rather, relaunched) its Xcelerator platform in June 2022 with a view to performing that function, as well as ticking both the ‘credible’ and ‘managed’ boxes that I refer to above. And Siemens is not alone, and comparable marketplace offerings are starting to become a fixture of the IoT space. But they could be much more. 

The key thing is their scale, able to offer a global platform for solution provider. For instance, Orange Business Services recently built a product in Australia called “Connected Sites”, which stemmed from a co-innovation initiative with Australian construction company McConnell Dowell. It is an end-to-end IoT solution focusing on construction and mining sites, based on a combination of the Microsoft Azure platform and LoRaWAN connectivity provided by partners. This is listed as a product on the Azure Marketplace (specifically, initial assessments are listed at a cost of USD9,000). Only a few companies offer the global scale that might allow Orange to sell these solutions out of footprint. And that’s before we even think about the familiarity that developers already have with the offerings from AWS, Microsoft etc. 

Differentiators: know thyself

Over the years I’ve seen a lot of companies attempt to find effective competitive differentiators in the tech space. There is little that is more pleasing than seeing an organisation work out what it has that others don’t and then boldly pursuing a strategy that exploits it. It may seem blindingly obvious but trust me it’s more unusual than you would think that tech organisations would take long hard looks at themselves to work out what they should do. 

I mention this because the final leg of my trip involved a visit to T-Mobile’s industry analyst event in Bellevue, Washington. Courtesy of the merger with Sprint, the new T-Mobile has a set of spectrum that is superior to its competitors in terms of delivering 5G, having decent holdings in all three of sub-GHz, mid-band and mmWave. T-Mobile saw an opportunity to establish clear blue water between itself and AT&T/Verizon, and it has gone for it. Now it’s no longer just the upstart competitor, it’s now also the flag bearer for 5G. And 5G permeates most of what it does, both consumer and enterprise. Its approach to the IoT arena, for instance, is very closely linked with 5G, most prominently in the new Advanced Industry Solutions. 

Of course, not all IoT is 5G, far, far from it. It’s not even going to be the dominant cellular technology any time soon, certainly if we just consider 5G New Radio. However, a lot of the market growth will be in 5G NR. 5G is where T-Mobile can stand out on IoT, so that is where it is focusing its attention.

There is also a brand value for T-Mobile in projecting a positioning based on technology leadership. The unconventional ‘Un-Carrier’ approach is still highly valid, but as a player with significant ambitions in the enterprise market it also needs to be something other than a commercial disruptor. Expanding on that to be a technological innovator is the next obvious step.

And now to write all of that up…

The above is just a little snapshot of some of the thoughts I've had on the state of IoT in the last few weeks. They will be distilled, expanded upon and wrestled into shape. Watch this space.