What is ‘Hyperscale IoT Connectivity’, and why do you need it?

Matt Hatton

By Matt Hatton, Founding Partner, Transforma Insights.

On the 26th February Transforma Insights published a White Paper entitled ‘Hyperscale IoT Connectivity could save enterprises 28% of their global connectivity cost’. In this article I explain how changing requirements for IoT necessitate a new approach to connectivity and how a new hyperscale approach can save enterprises over $100 billion in operational costs.

The starting point is the current state of IoT connectivity. A decade ago, the ways in which global connected devices were supported was relatively simple. You put a SIM card into a 2G/GPRS modem and the device could go almost anywhere in the world and, courtesy of long-established roaming agreements, could connect on at least one network, and data flowed via the public internet to the client. Simple, but not very flexible or cheap. In the intervening time a much richer set of connectivity tools has come available, including new network technologies such as 5G, LPWA and eSIM. The network architecture has also changed radically with the growth of cloud services, edge computing and the arrival of the hyperscalers. The need for regulatory compliance has also grown exponentially. To fully take advantage of the new technologies available and to mitigate the risk associated with the new commercial and regulatory reality, connectivity offerings must also change.

For this reason, we advocate for a connectivity solution with a number of particular characteristics that make it more suitable for IoT in the 2020s. Firstly they must be local, which includes localising connectivity using eSIM/eUICC, as well as data management; local compliance is the key. It must also be adaptive to different architectures including cloud and edge, as well as quick to deploy into new territories. Thirdly it needs to be secure, extending the enterprise’s security perimeter out to the edge device. Finally, it must be collaborative and interoperable between the diverse members of the IoT ecosystem. We term solutions that meet these criteria ‘Hyperscale IoT Connectivity’.

The report also outlines USD117 billion in cost savings that can be made by enterprises over the next decade by using this Hyperscale approach. This won’t necessarily mean a lower sticker price for data plans. More important than that, it substantially reduces many other costs associated with connecting devices, streamlining processes and accelerating time-to-market. Compared to these substantial operational costs, the spend on connectivity is very moderate. The cost of choosing poorly can make a substantial difference to the profitability of a product or service.

In total we identify eleven main ways in which enterprises will benefit from selecting a Hyperscale-friendly solution. These range from greater supply chain efficiency, courtesy of using eSIM, through to more streamlined device-to-cloud integration giving a superior and more consistent global deployment experience. Other benefits include faster time-to-market for companies deciding the move into overseas markets, and superior security mitigating the cost of layering on additional measures. Further details can be found in the White Paper.

chart: additional spend and mitigation of costs as proportion of global cellular IoT connectivity revenue 2020-2030

On average, our analysis tells us that enterprises can save costs equivalent to 28% of their connectivity spend. This figure hides a lot of variation. In the manufacturing vertical, for instance, the cost savings over the next decade could more than exceed the total spent by the sector on connectivity.

The last ten years has handed us a wide range of powerful tools to better connect IoT devices. What is required to really harness them is a product offering that both reflects those technical capabilities, and the new commercial realities of how IoT is deployed in the 2020s.

The White Paper is available to download here, including more details on all of the topics discussed in this article, as well as a description of the methodology used.

Related posts