Good ideas and bad ideas for Communications Service Providers in IoT

Communications Service Providers are faced with choosing between a plethora of different strategies for addressing the IoT market, reflecting the rapidly evolving technical and commercial landscape. In this article I present a view of what strategies are likely to be the best for most MNOs/MVNOs related to IoT.

In July 2024, Transforma Insights published its annual CSP IoT Peer Benchmarking report which examined the strategies and capabilities of 25 leading MNOs and MVNOs in addressing the IoT opportunity. Based on that research – and the ongoing work we do at Transforma Insights analysing the key themes impacting IoT – we can identify a long list of major trends in the IoT connectivity landscape. The key thing for the CSPs themselves is to understand which of those trends are the most critical, which are outliers and which are dead-ends.

The first of the trends identified in the report is that the pace of change is rapid. Possibly more rapid than it has been for many years. New technologies, notably SGP.32, continue to surface with potentially disruptive implications. At the same time new commercial models are coming to the fore. Just think of Vodafone’s newly spun out IoT business, Softbank’s acquisition of Cubic Telecom, or Verizon’s Global IoT Orchestration; to name but a few of the major new developments in the last year. There have also been some more gradual trends, for instance towards a greater focus on devices and consulting. With this rapid pace of change, CSPs need to be continually reviewing their approach. ?

A word of warning about this article. The act of proposing a set of approaches that are better or worse is inherently reductive. What works for one operator may not work for another, and vice versa. Every company finds itself in a different starting position and the approach to addressing the market will necessarily vary. Indeed there are advantages to doing so: it’s often good to zig when the rest of the world zags. For instance, for some operators a low-touch connectivity-only approach making good use of online channels will be appropriate. For others, with a more consultative selling approach, the focus will be rather different. There’s room in the market for both types of players. Another good example is moving ‘up-the-stack’ into vertical solutions. We generally characterise this as being a ‘bad idea’ but it very much depends on the market that the company operates in, the vertical being addressed, its other assets, and its overall right to play. The answer for a specific MNO/MVNO will be different. If you want to know what’s best for your company, the best thing to do is speak with us.

Sounds like a good idea, is a good idea

Let’s start with the ‘good ideas’. I’ve tried to steer clear of anything which is too blatantly obviously a good idea, and therefore already universally done. An example would be having a dedicated set of OSS/BSS (in the form of a Connectivity Management Platform) for IoT. Any company that is in any way serious about IoT connectivity already has that, therefore such a recommendation would be unnecessary. Instead I have focused on those where perhaps MNOs/MVNOs have been a slow to react or have otherwise missed a trick.

Other than the recommendation that any MNO/MVNO should work with Transforma Insights (naturally), several of these most obvious good ideas relate to the breadth of the commercial proposition. Most CSPs are coming around to the idea that the IoT connectivity proposition extends in one direction to the cloud (i.e. with cloud connectors and developer tools) and in the other direction to the device. This has manifest itself in more consulting focused on device/connectivity cross-optimisation and generally helping the customer navigate their way through building an solution. Additionally it has come to the fore as holistic device/connectivity twins (integrating connectivity management and device management functions) for troubleshooting across both elements of the proposition. It is also increasingly taking the form of more refined hardware propositions including extending out to sensors and including management of gateways (and sensors in some cases). Examples here include Telenor’s IoT Complete offering, KORE’s OmniHub and Eseye’s Hera router. We have been saying for a while that really IoT connectivity providers are ‘device-to-cloud’ providers and that is increasingly reflected in the propositions.

On the subject of consulting, we should also note that ‘compliance-as-a-service’ is an increasingly important element of a connectivity proposition. The regulations related to IoT (particularly security) have become increasingly strict and there’s lots more to come. Any company deploying IoT needs help understanding their compliance requirements. Check out our regulatory database for more details.

The other aspect of enhancing the commercial proposition is via ‘contextualisation’. This is a drum that Transforma Insights has been beating for several years: the positioning of what is ostensibly a horizontal offering (i.e. IoT connectivity) to be as relevant and contextualised as possible for the vertical into which it might be deployed. This partly takes the form of ‘talking the talk’ of the vertical concerned. It also partly takes the form of modular and composable micro-services (including devices, VAS, networks, pricing components and so on) that can be structured to address the specific requirements of a particular vertical in a relatively simple way.

Next after that are a few areas associated with streamlining operations. This includes the use of cloud-based infrastructure and MNOs acting more like MVNOs (i.e. being more nimble and reactive). We’ve seen both approaches surface quite a lot in the last few years. Both are good ideas to an extent but not overwhelmingly so. Hosting everything in the cloud isn’t necessarily more efficient, although it can be and can make for easier compliance mechanisms. And imitating an MVNO is only a successful strategy if it really means more streamlined processes. So treat both with a little caution.

What has been conspicuously lacking from this discussion of good ideas thus far has been technology. Release 17 NTN satellite connectivity is going to be a valuable addition to a proposition but more as a hybrid/fallback/overage capability and it is likely to be a few years before it makes a substantial impact on the market. We have seen good work from DT IoT and Emnify here recently. Certainly useful, but probably doesn’t move the needle too much in the immediate future.

Sounds like a bad idea, actually a good idea

There are also several network technology options that fall into the ‘good idea but sounds like a bad idea’ quadrant. LTE Cat 1bis is one example. With some of the challenges of NB-IoT/LTE-M (more on which later) it is perhaps unsurprising that there is rightly some appeal to an LTE variant which has the advantage of near-universal availability (albeit with inferior in-building coverage and battery life). The most prominent, however, is the new ‘IoT’ remote SIM provisioning standard SGP.32. Just a year ago there was quite a lot of hesitancy from the MNOs (although not the MVNOs) about the likely disruption that SGP.32 might cause. Today, correctly, most now see it as more of an opportunity than a risk and we think that the concerns about SGP.32 heralding a dramatically more footloose subscriber base are largely unfounded. We even see it triggering more inter-MNO collaboration.

Another couple of themes within the same quadrant relate to Connectivity Management Platforms. This had been quite a settled space for many years, dominated by Cisco, Ericsson (now Aeris) and a few in house platforms such as Vodafone’s GDSP and Verizon’s ThingSpace. Some commentators would suggest that the optimum approach is a single CMP, but the reality is that we see quite a lot of value in multi-sourcing (i.e. having more than one platform which can be used for different use cases and revenue profiles) and in implementing an abstraction layer above the CMP to allow for easier integration of multiple platforms (either your own or those of third party partners). Several major operators have successfully implemented a multi-platform approach, most obviously Deutsche Telekom and Telefonica. We expect more to follow.

Another thing that we consider generally a ‘good thing’ despite the bad press, is network attach fees (or whatever your preferred terminology might be). Historically IoT offerings piggybacked on networks deployed to overwhelmingly support handset users. However, with the advent of NB-IoT/LTE-M, MNOs found themselves having to invest substantial amounts in network infrastructure/upgrades that were focused entirely on IoT. It’s hardly surprising that MNOs want to recoup the cost of running those networks, and attach fees (annoying though they might be) are really the most appropriate way of doing that. I should note, that if you’re an MVNO you certainly won’t think of these as a good thing!

Finally in the quadrant, Fixed-Wireless Access. Quite a few of the CSPs profiled in this report and most prominently the MVNOs (and others not profiled), have focused attention on the Fixed Wireless Access (FWA) opportunity in the last year or so. This is particularly pronounced in the US. The appeal is obvious, with much higher revenue per connection and potentially greatly benefiting from the ability to deliver multi-network support. The offerings are often as an enterprise managed gateway targeted at particular verticals with deployment and managed sensors.

Sounds like a good idea, actually a bad idea

It was a moot point whether to tackle the bottom right quadrant next. This is where the approaches that are likely to be popular but aren’t (we think) a good idea sit. And I should reiterate that this doesn’t apply to everyone. For some, some of these things will be a good approach.

Let’s start with channels. Online channels are still mostly seen as a lead generation exercise, particularly given that most CSPs will favour a more consultative selling approach as discussed in the previous section. There are exceptions, notably 1NCE where online channels is a key focus and very well aligned with its overall proposition. For many, however, such as Eseye and Velos IoT, online channels are not seen as particularly compatible with their approach of supporting complex and customised projects. At risk of contradicting myself, I would say further in the defence of online channels that they can be a good way to demonstrate a joined-up fulfilment process.

Anyone who is anyone is using AI as a competitive differentiator today, right? In IoT connectivity not really. In a recent report, Transforma Insights presented the ways in which AI was being used by the CSPs profiled in this report (see ‘How are MNOs and MVNOs harnessing AI for their IoT operations?’ (June, 2024)). Generally, AI is used for relatively mundane use cases that are not really specific to the provision of IoT connectivity, for instance in administrative tasks (e.g. invoice analysis) or coding. However, some are making use for anomaly detection, network management and commercial planning. Of course some of those CSPs that provide high end consulting/SI are also involved in combining IoT and AI, but in terms of making use of AI as a differentiator in their own operations, we see no concrete examples where that is really the case.

I’ve mentioned the idea of moving ‘up the stack’ already. Interest in delivering vertical solutions, such as fleet management or track & trace, waxes and wanes over the years. The appeal is obvious, and many commentators will offer the view that the best opportunity to make money in IoT is to move out of connectivity (which accounts for maybe 5-10% of revenue) and into the full solution (which is most of the rest). A simple suggestion, but usually wrong. Generally at Transforma Insights we urge caution when it comes to building full solutions because such CSP offerings tend to be very ‘me too’ and launched into markets that are highly competitive. CSPs will only succeed where they have a right to play, for instance with some differentiator in terms of channels, where the connectivity aspect of the solution is a critical part, or where there is a nascent market with limited incumbent players.

The other subjects in this quadrant relate to LPWA technologies: NB-IoT, LTE-M and 5G RedCap. Taking the latter first, we don’t see RedCap as being currently mature enough. Our view has not changed since July 2022 (‘What is 5G RedCap and how does it fit into the portfolio of cellular IoT connectivity technologies?’) when we identified that RedCap as it currently stands is optimised for only a limited range of use cases. Further evolutions will likely make it more appropriate for IoT, but that’s a few years off yet. Turning to LTE-M/NB-IoT, Transforma Insights regularly engages with enterprises wishing to identify vendors to support their IoT deployments. In several cases this has involved the use of NB-IoT and LTE-M. Such deployments continue to be dogged by challenges. In some cases this relates to the availability of networks, with some countries remaining blank spots or with limited coverage, making reliance on these LPWA technologies challenging, particularly for deployments across multiple territories. In other cases there are persisting commercial issues with some operators prohibiting roaming or presiding over unreliable interconnect between networks. The experience is not currently great for multi-country deployments, although single country is typically fine. MNOs/MVNOs should not be (a) making things worse by putting limits on roaming or (b) encouraging reliance on these technologies without a fallback.

Sounds like a bad idea, is a bad idea

Now we get into the category of varying degrees of ‘I can’t believe companies are (still) doing this’. The least bad example is public LoRaWAN. It’s a great technology, but one where private deployments have generally proved to be the most appropriate. For MNOs, hardly anyone loves their LoRaWAN deployment, although for supporting some applications (smart water metering being the obvious one) it can still be quite useful. While we’re on the subject of networks, lets address 4G switch-off. I thought I’d be long retired before anyone wanted to turn off LTE, but there’s already some talk about it. Please don’t. It will remove the one reliably globally deployed network technology. At least for the next decade.

The next category is one that also needs a caveat or two: competing solely on price. There is room in the market for price competitors, as there is with every market. But at this point connectivity prices are really quite low and even vendors with a heavy value focus need other assets to bring to the offering.

Significant strides have been made in recent years in using more appropriate mechanisms for supporting connectivity across multiple geographical markets. This includes both on compliance with permanent roaming rules (and we note in the report the great progress made on, for instance, being compliant in Brazil) and not trying to exploit arbitrage arrangements for consumer roaming and/or sponsored roaming. But there are still a few that sit in a grey area of compliance, either with regulations or with the rules of the host operators. Don’t do this.

Last but not least, competing with your customers. The IoT connectivity space has a lot of overlapping roles. MVNEs are also MVNOs, hardware vendors sell direct and through other MNOs/MVNOs, platform vendors are also MVNOs. As such, there is some inevitability about a touch of confusion about being a supplier at the same time as being (usually only nominally) a competitor. But don’t let that fool you. Everyone in the space is still very much conscious of what other players in the value chain are doing and generally won’t take kindly to someone impinging on their space.

Learn more

If the above discussion has been interesting, this is just a small sub-set of the discussion that we include in the Communications Service Provider IoT Peer Benchmarking report and more broadly in the Transforma Insights Advisory Service. Contact us for more details.

?