O2 IoT: The Invisible IoT Giant

The 16.4 million connection contradiction

Ask a UK business to name a mobile operator with a serious Internet of Things business and Vodafone will usually come first. O2 may not come up at all. That is odd, because O2 sits on one of the largest connected-device estates in the country.

In its results for the second quarter of 2026, covering the three months to the end of June, Virgin Media O2 reported that its IoT estate had grown to around 16.4 million connections, up from 15.4 million a year earlier. Against a total reported mobile base of roughly 46 million connections, that means somewhere close to one in three connections carried on the O2 network is now a machine rather than a person.

Now look at how O2 presents itself to business. In 2026 the operator relaunched its entire business arm under the name O2 Business, built on the merger of Virgin Media O2 Business and Daisy Group. The new brand is organised around connectivity, mobile, unified communications and simplifying technology for UK organisations. IoT is not a headline category. Smart Connect, its IoT SIM management platform, is not front and centre. Managed IoT connectivity is not one of the propositions a first-time visitor would trip over.

So you have an operator where a third of its mobile connections are IoT, and a business website where IoT is close to invisible. That gap is the story. It is not that O2 missed IoT. It is that O2 built a very large IoT business and then, for reasons that are partly deliberate and partly structural, chose not to talk about it.

Does O2 actually have an IoT business?

Yes. A large one. The confusion comes from expecting an IoT business to look like Vodafone's, with a named global platform, sector pages, device catalogues and a clear route from a web search to a sales enquiry. O2's IoT business does not look like that. It is spread across several different things that are usually discussed separately, and it includes a great deal of infrastructure and wholesale that never carries an O2 badge in front of the end customer.

The evidence that O2 has genuine IoT capability is not thin. It has delivered one of the largest machine-to-machine contracts ever signed anywhere. It connected Tesla vehicles across Europe. It was the first UK operator to launch a national LTE-M network for low-power IoT. It has built private mobile networks for factories, hospitals and farms. It sits behind a large share of the multi-network and roaming SIMs that UK IoT aggregators sell. And it has access to Telefonica's international IoT platform.

The right question is therefore not whether O2 has an IoT business. It is why that business is so hard to see, and what that means for anyone trying to buy IoT connectivity in the UK. To answer that, it helps to separate the different things people lump together when they say O2 IoT.

It is worth being precise about what makes an operator a serious IoT provider, rather than simply a network that some IoT devices happen to use. A serious IoT operator runs connection-management tooling built for machines rather than phones, supports low-power technologies for battery devices, handles long device lifecycles measured in years, manages roaming and multi-network behaviour, and can operate regulated, mission-critical estates without failure. By every one of those measures, O2 qualifies. It has run a national, long-life, regulated IoT programme for over a decade. It launched low-power connectivity ahead of its rivals. It manages large fleets through a mature platform. The capability test is not close. O2 passes it comfortably. The visibility test is where it falls down.

The five layers of O2 IoT

Most confusion about O2 and IoT comes from treating it as one product. It is not. There are at least five distinct layers, and a single connected device can touch several of them. Crucially, the layer that carries the traffic is often not the layer the customer sees.

Behind all five sits a sixth element that rarely surfaces in the UK conversation: Telefonica's international IoT operation and its Kite management platform. O2 is the UK arm of Telefonica, so in principle a UK deployment can reach into a global footprint. In practice, the public material does not join those dots for a buyer.

Walking the layers, one by one

The radio network is the foundation. O2 operates a national mobile network across 2G, 4G and 5G, and it added LTE-M in 2020 for low-power devices. Any IoT device that connects over O2 is using this layer, whether or not the customer has any commercial relationship with O2. A smart meter, a vehicle telematics unit or a tracker bought through a third party can all sit on O2 radio while the customer never types the word O2.

Direct connectivity is the layer most people picture when they think of buying from an operator: an O2-branded business SIM with a data plan, sold and supported by O2 Business. This is the visible retail face of O2 IoT, and it is real, but it is only a slice of the total.

Managed connectivity is where Smart Connect lives. A raw SIM is not much use across a fleet of thousands of devices without a way to activate, suspend, monitor, diagnose and bill them at scale. Smart Connect is O2's answer to that, and it is the difference between selling data SIMs and running an IoT connectivity business.

Wholesale and partner connectivity is the largest and least visible layer. O2 sells network access to mobile virtual network operators, IoT aggregators, multi-network SIM providers and resellers, who then build their own propositions on top. The end customer buys from the partner and may never know O2 is underneath.

Private networks and integrated solutions are the newest and most heavily marketed layer. Rather than connecting a device to the public network, O2 builds a dedicated 4G or 5G network on a customer's own site, often with a partner such as Nokia, and often as part of a wider digital-transformation project. This is prestigious, bespoke work, and it is where most of O2's recent IoT publicity has gone.

Understanding these layers separately is the single most useful thing a buyer can do, because a proposal that sounds like one thing may actually be another. A quote for a private network is not a quote for managed public IoT connectivity. A multi-network SIM that happens to use O2 is not the same as an O2 contract. The layers behave differently on price, support, coverage and control.

Reading the 16.4 million honestly

Sixteen point four million is a striking number, and it is real. But a good reference should say what a number contains before leaning on it. Two points matter here.

Some of the growth is a consolidation, not organic activation

Virgin Media O2 reported the year-earlier figure for the second quarter of 2025 as 15.4 million. Trade coverage at the time actually reported a lower figure of around 13.2 million for that same quarter. The difference is not an error. From the start of 2025, Virgin Media O2 began consolidating Daisy Group into its reporting, and the IoT base was restated accordingly. So part of the apparent jump reflects Daisy's connections being folded into the count, not a surge of newly activated devices. Anyone quoting a clean multi-year growth rate for O2 IoT needs to be careful: comparing a pre-Daisy figure with a post-consolidation figure overstates organic growth.

A large slug of the base is smart metering

The single biggest driver of O2's IoT count is not a fleet of business customers buying SIMs one deployment at a time. It is national infrastructure. O2 connects a very large share of Great Britain's smart meters under a contract that predates most of the current IoT market. Those meters sit quietly in the background. They do not generate reviews, router bundles, distributor announcements or dozens of separate case studies. They generate a large connection count and very little visibility.

What is probably inside the number

Because Virgin Media O2 does not publish a full breakdown, the composition of the 16.4 million can only be estimated from what is publicly known, and it should be read as informed inference rather than disclosed fact. The largest single block is almost certainly smart metering, given the scale of the national contract. Beyond that, the likely components include automotive and vehicle telematics, security and alarm systems, asset and vehicle tracking, utilities and metering beyond electricity and gas, payment terminals, connected consumer products, direct enterprise deployments across sectors, and a substantial volume of wholesale partner connections.

The important question a reference reader should hold onto is not the exact split. It is whether these 16.4 million connections are mostly commercially owned, directly served O2 customers, or whether a large share sit inside major infrastructure programmes and wholesale relationships where O2 is the network but not the customer-facing brand. On the balance of the evidence, it is the latter. That is not a criticism. It is simply what the number is.

How O2 got here: a timeline

The scale of O2's IoT estate is not an accident of the last two years. It was built over more than a decade, through a small number of very large decisions. The problem is that most of the loudest moments are now buried in old press releases.

Read as a list, this is the record of a serious IoT operator. Read as a marketing story, it is a series of disconnected moments, each celebrated once and then left to gather dust. The four biggest are worth taking in turn.

The smart-meter contract

In 2013, Telefonica UK, which operates the O2 network, signed two 15-year agreements to provide the communications infrastructure connecting smart meters across the Central and Southern regions of Great Britain. The combined value was around 1.5 billion pounds. The northern regions went to a consortium led by Arqiva under a separate, smaller contract. The solution was based primarily on O2's cellular network, supplemented by mesh technology for meters in hard-to-reach locations. O2 described it as the largest machine-to-machine contract of its kind at the time, ultimately underpinning connectivity at more than 20 million premises.

This single contract disproves any notion that O2 has never been serious about IoT. It is national, long-life, highly regulated infrastructure, exactly the kind of deployment that separates a genuine IoT operator from a reseller of data SIMs.

Tesla and the connected car

In 2014, Telefonica announced it would provide machine-to-machine connectivity for the Tesla Model S across several European markets, with UK connectivity running over O2. The service supported navigation, in-car media, internet access and remote vehicle diagnostics, and was managed through Telefonica's global machine-to-machine platform. The agreement established Telefonica as Tesla's largest European connectivity partner at the time. It is a significant connected-car reference, and it has all but vanished from O2's current IoT identity.

The LTE-M launch that went quiet

In February 2020, O2 announced it would become the first UK operator to introduce a national LTE-M network, the low-power cellular technology suited to asset tracking, sensors, alarms and metering. It began with 50 live sites and reached around 10,000 sites by the middle of the year, with coverage described at the time as reaching well over half of premises and population, and a national rollout planned to complete that year.

Then the public story largely stopped. O2 still has pages referring to LTE-M and to NB-IoT, but there is little recent public detail on current coverage, on which tariffs support LTE-M, on whether wholesale partners and inbound roaming SIMs can use it, or on the present state of NB-IoT. The contrast with the confident 2020 launch is stark, and it is the pattern of this whole article in miniature: a capability announced loudly, then left largely unspoken.

Private networks for industry

The most visible recent O2 IoT work is in private networks. In 2022, Virgin Media O2 Business and British Sugar switched on a multi-site private 4G network spanning four factory sites across Norfolk, Suffolk and Nottinghamshire, built with Nokia, to support automation, robotics, predictive maintenance and even health-and-safety drones. The same year, Virgin Media O2 Business switched on what it described as the UK's first 5G-connected hospital with South London and Maudsley NHS Foundation Trust, running clinical observations, smart medicine storage, air-quality monitoring and augmented-reality remote support over a private 5G network. In 2025, a portable private 5G network was trialled at Overbury Farms, a large arable and livestock estate, connecting AI pest traps, water-quality sensors and environmental monitoring as part of a government-backed rural wireless programme.

These are strong case studies. But they are private-network stories, not evidence of a strongly marketed public IoT SIM service. That distinction matters for a buyer, and we come back to it below.

Why private networks do not fill the public IoT gap

The British Sugar, hospital and Overbury projects are genuinely impressive, and they show that O2 can integrate complex IoT deployments end to end. But private-network IoT and public managed IoT connectivity are different markets that happen to share the letters IoT. A private network is a bespoke, site-specific build, usually a large capital project delivered with partners, aimed at a single customer's premises. Public IoT connectivity is a volume product: SIMs and connection management sold to many customers to connect devices that roam across a shared national network.

O2 is visibly good at the first and quiet about the second. That is the wrong way round for building a broad IoT reputation, because the second market is where most IoT buyers actually live. A manufacturer that wants to connect a thousand devices in the field does not need a private 5G network at a factory. It needs a clear, well-supported public IoT connectivity proposition it can buy and manage. That is precisely the thing O2 is least visible about.

The smart-meter foundation

To understand why O2 looks smaller in IoT than it is, start with smart meters. A single national programme can put millions of connections onto one operator's network without producing any of the noise a normal product business generates.

Those meters do not appear in comparison articles. They do not generate router bundles or distributor promotions. They do not create thousands of visible customer reviews. They sit behind infrastructure, connecting quietly for fifteen years at a time. So a large chunk of O2's IoT count is real, load-bearing, and almost completely invisible to the ordinary IoT market.

There is also a live technical wrinkle worth flagging. The original smart-meter network was built on older cellular generations. As UK operators retire 2G and 3G, the smart-meter estate faces an upgrade path towards 4G, and that work is now underway across the industry, including trials involving other operators. In other words, even O2's single largest IoT trophy is not a static asset. It is a fleet of connections that has to be modernised, and the modernisation is not a purely O2 story. That is a useful reminder that in UK IoT, the network on the SIM and the network doing the work are not always the same thing.

The lesson infrastructure IoT teaches

The smart-meter story carries a wider lesson about reading any operator's IoT numbers. Infrastructure IoT, the meters, the national programmes, the long-life fixed estates, inflates connection counts massively while contributing almost nothing to public visibility or to the kind of self-service revenue that markets to ordinary buyers. An operator can therefore look like a giant on a results slide and a ghost in the market at the same time, and both impressions can be accurate. When you see a headline IoT connection figure from any UK operator, the first question worth asking is how much of it is a handful of very large infrastructure contracts, and how much is a broad base of served customers. For O2, the honest answer is that a great deal of the headline is infrastructure, which is exactly why the headline and the market presence feel so mismatched.

O2 Smart Connect explained

Smart Connect is O2's IoT SIM management proposition. It originated with the operator's 2012 partnership with Jasper Wireless, the connection-management platform later acquired by Cisco and rebranded as Cisco IoT Control Center, which now sits behind many operators worldwide. O2 has not published a clear current statement of what Smart Connect runs on today, and it has announced no change, so it is fair to describe the proposition by what it does rather than by its internals.

On O2's own description, Smart Connect covers the core of connection management that any serious IoT deployment needs: activating, suspending and deactivating SIMs, real-time diagnostics, alerts and rules, tariff and billing control, and access through APIs. It offers both physical SIMs and embedded options. It describes access to other UK networks where required, and it references a large number of roaming agreements and a link into Telefonica's global footprint.

That last point is more interesting than it first looks. O2 does not present Smart Connect as a single-network, O2-only service. Its own material describes the ability to use alternative UK networks where O2 is unavailable. For an IoT buyer, that raises a set of practical questions that O2's pages do not fully answer.

Why the platform question matters

It might seem pedantic to ask what Smart Connect runs on. It is not. The connection-management platform behind an IoT SIM determines the quality of the tools a customer gets: how granular the usage alerts are, how good the diagnostics are when a device stops reporting, how flexible the tariff and rate-plan controls are, how mature the APIs are for automating a large fleet, and how the platform handles international profiles. Two SIMs that both claim SIM management can offer very different day-to-day experiences depending on the platform underneath.

O2's heritage here is strong. The 2012 Jasper Wireless deployment put it on one of the most widely used connection-management platforms in the industry, the lineage that became Cisco IoT Control Center. That is a genuine advantage. The weakness is presentational: because O2 does not clearly state the current platform, its international reach or its roadmap, a prospective customer cannot easily compare Smart Connect against a rival platform, or against Telefonica Kite, on the things that actually differ. The capability may be excellent. The shop window makes it hard to judge.

Smart Connect versus Telefonica Kite

Behind O2 sits Telefonica's Kite platform, a mature international IoT connectivity and management proposition. On its public description, Kite spans global SIM connectivity across 2G to 5G, LTE-M and NB-IoT, real-time usage control, analytics dashboards, SIM location, diagnostics, security and anomaly detection, APIs and device management. It is plainly not a neglected sideline; it is the core of a large international IoT operation.

The awkwardness for a UK buyer is that O2 does not clearly connect the two. It is not obvious from O2's UK-facing material whether a British customer can buy Kite through O2, whether Smart Connect is the UK alternative to Kite, whether the two serve different customer sizes, or how a UK deployment would expand internationally using Telefonica's footprint. The capability exists; the map to it does not.

The bottom two rows are the point. The gap is not capability. It is articulation. A buyer cannot easily tell where Smart Connect stops and Kite starts, or which one they are actually buying.

The international question O2 does not answer

For any business planning to deploy IoT beyond the UK, the Smart Connect and Kite relationship is not academic. It determines who they contract with, who bills them, whose profiles their devices use abroad, and who supports them when something breaks in another country. Telefonica's international footprint is a real asset, and in principle it should make O2 a strong choice for a UK business that expects to expand overseas. In practice, the absence of a clear public explanation means a buyer cannot tell whether that footprint is available to them, on what terms, and through which contracting party. Vodafone answers this question directly with a single named global IoT proposition. O2 leaves it for a sales conversation, which is friction at exactly the moment a buyer is trying to compare options. For international IoT in particular, that lack of a clear answer is a competitive disadvantage that has nothing to do with the quality of the underlying network.

LTE-M, NB-IoT and the LPWAN question

Low-power wide-area networking is where the UK operators made visibly different bets, and it is a useful lens for anyone trying to understand the market rather than just O2. LTE-M and NB-IoT are both standardised low-power cellular technologies, but they suit different jobs. LTE-M offers higher throughput, lower latency, mobility and handover, and support for voice, which makes it suitable for asset tracking, alarms, medical devices and moving assets. NB-IoT is optimised for very low bandwidth, deep indoor coverage and long battery life on largely static sensors.

How LTE-M and NB-IoT actually differ

Both technologies are part of the 3GPP low-power wide-area family, and both are designed for devices that send small amounts of data, run for years on a battery and reach places ordinary mobile data struggles with. The differences decide which one suits a given job.

LTE-M, sometimes written Cat-M1, uses a wider channel and offers meaningfully higher throughput than NB-IoT, along with lower latency, support for mobility and handover between cells, and support for voice. Those properties make it a good fit for anything that moves or needs to respond quickly: vehicle and asset tracking, alarms and security devices, wearable and medical devices, and any application where a device roams across the network rather than sitting still. LTE-M devices can also make use of power-saving features such as Power Saving Mode and extended Discontinuous Reception, which let a module sleep deeply between transmissions to preserve battery life.

NB-IoT trades throughput for reach and endurance. It uses a very narrow channel, which gives it excellent deep-indoor and below-ground penetration, the ability to support very large numbers of devices per cell, and extremely low power draw. It suits static, low-bandwidth sensors: utility meters in basements, environmental and agricultural sensors, parking bays, and fixed monitoring equipment. Its weaknesses are the flip side of its strengths: limited throughput, higher latency, and historically weaker support for mobility and voice.

A useful rule of thumb is that LTE-M suits things that move or need responsiveness, and NB-IoT suits things that stay put and need to last. Many large deployments end up using both, matched to the device. This is exactly why an operator's LPWAN choice matters to a buyer, and why the UK split, with O2 leading on LTE-M and Vodafone on NB-IoT, shapes which operator is the natural fit for a given fleet.

It also explains a subtle procurement trap. A module certified and optimised for one operator's low-power network is not automatically the right choice for another's. Coverage, band support, roaming behaviour and the specific power-saving configurations an operator enables all vary. For a fleet that must work across multiple UK networks or internationally, this is one of the strongest arguments for a multi-network SIM or, increasingly, for remote SIM provisioning under SGP.32, so that the network identity is not fixed in the factory.

O2 went first on LTE-M. Vodafone built its low-power strategy around NB-IoT, deploying it across its European footprint. Three trialled its own LTE-M approach. That split is why a device certified for one operator's low-power network is not automatically a fit for another's, and it is a large part of why LPWAN procurement in the UK is more fiddly than it should be.

For a practitioner, the practical takeaway is simple. Do not choose an LTE-M or NB-IoT device on the strength of a launch announcement. Confirm the current, contracted position for your exact sites, your exact modules and your exact roaming needs, because the public picture is several years stale.

O2 as a wholesale network

This is where an understanding of the market adds the most value, and where the invisibility starts to make commercial sense. A very large amount of O2's IoT traffic reaches the end customer through somebody else.

O2 holds a substantial wholesale position. Its network is reached through MVNOs, business communications providers, IoT aggregators, multi-network SIM suppliers, connectivity resellers, white-label providers and roaming agreements. An IoT customer can use O2 extensively without ever buying an O2-branded service. The commercial relationship, the management portal, the first-line support, the tariff design, the hardware recommendation and the billing all belong to the intermediary. O2 earns wholesale revenue and stays in the background.

This is also why the phrase multi-network matters so much in UK IoT, and why it no longer means what it once did. A multi-network or multi-IMSI SIM is designed to select the strongest available UK network rather than lock to one. In practice, in the UK that now means the three mobile network groups: Vodafone, O2 and EE, following the consolidation of the old four-operator market. So a business buying a multi-network IoT SIM through an aggregator is very likely using O2 as one of its underlying networks, alongside Vodafone and EE, without ever appearing on O2's books as an IoT customer.

For O2, this is a perfectly rational position. It can sell network capacity at scale without building the sales, support and specialist product organisation required to serve every 50, 500 or 5,000-SIM customer directly. The risk is equally clear: the aggregator becomes the recognised IoT brand, and O2 becomes an interchangeable layer underneath. That is a comfortable place to earn money and a poor place to build a reputation.

What the wholesale chain looks like in practice

It helps to trace the chain for a typical multi-network IoT SIM. At the bottom sits the radio network, which for a UK deployment may be O2, Vodafone or EE, sometimes more than one. Above that sits the operator whose profile is on the SIM, which may be one of those networks or a specialist connectivity provider using their infrastructure. Above that sits the provider of the management portal the customer logs into. Above that sits the reseller who designs the tariff and issues the bill. Above that sits whoever provides first-line support. And at the very top sits the brand the customer believes they are buying from.

Any of these can be a different company. A single connected sensor might use O2 radio, carry a specialist provider's profile, appear in an aggregator's portal, be billed by a reseller and be supported by an integrator, with O2 invisible at every visible step. This is not unusual. It is how a large part of the UK IoT market actually works, and it is the practical reason O2 can be enormous in IoT while feeling absent.

What it means if you are buying

For a buyer, the lesson is to look past the badge on the SIM and ask what actually underpins the service. If resilience matters, confirm which UK networks a multi-network SIM can reach and how it switches between them. If coverage at a specific site matters, test on the actual networks in use, not on a brand. If support matters, establish who answers the phone when a device goes dark, because it may not be the operator whose network is carrying the data. And if international expansion is on the horizon, establish early who owns the relationship once a device leaves the UK, because that is where the O2, Telefonica and partner boundaries become real. The right choice is rarely about which brand is most famous. It is about which combination of network, platform and support fits the fleet.

Why Vodafone owns the public IoT narrative

Vodafone's advantage over O2 in IoT is not mainly about network technology. It is about packaging. Vodafone built Vodafone IoT as a recognisable international business with a global connectivity narrative, a named managed platform, integrated device propositions, automotive heritage, clear sector pages and a fairly direct route from research to a sales enquiry. O2 has comparable individual components, but they are divided between O2, Virgin Media O2, O2 Business, Telefonica and a web of partners.

The difference shows up most clearly in the answers a buyer receives to a few basic questions.

That final row is the heart of it. Vodafone built an IoT brand. O2 built a collection of capabilities. Both can carry your traffic. Only one makes it easy to understand what you are buying.

This is a commercial gap, not just a cosmetic one

It would be easy to dismiss all this as marketing polish, but the gap has real commercial consequences. When a buyer researches IoT connectivity and finds a clear, confident, well-documented proposition on one side and a scattered set of pages and press releases on the other, the clear proposition wins consideration before a single conversation happens. Procurement shortlists are built from what can be understood quickly. An operator that makes a buyer work to understand what it offers is an operator that gets left off shortlists, regardless of how good the underlying network is.

There is also a self-reinforcing effect. Because O2 markets IoT lightly, the specialist aggregators and integrators fill the gap, become the recognised faces of IoT connectivity, and capture the customer relationships. That makes it rational for O2 to keep leaning on wholesale, which makes it market IoT even more lightly. The invisibility becomes structural. Breaking out of it would require a deliberate decision to compete for the visible customer relationship, not just the wholesale revenue underneath it.

O2 and SGP.32: the next test

The clearest current test of whether O2 will change is SGP.32, the GSMA specification for remote SIM provisioning built specifically for IoT. SGP.32 lets an operator or platform add, switch and remove operator profiles on a device over the air, across a fleet, without physical access and without the constraints of the older M2M and consumer eSIM standards. The certification baseline is the SGP.32 version 1.2 specification, with a version 1.3 published in 2026; deployments today are built against the version 1.2 baseline.

Through 2025 and into 2026, SGP.32 moved from specification into commercial deployment. The first movers to reach live delivery have been the likes of Telenor IoT, alongside eSIM management platforms and a growing set of certified hardware. The UK network groups, EE, the combined Vodafone and Three business, and Virgin Media O2, are each described as building SGP.32-native offerings on top of their networks, at various stages of readiness.

Here is the pattern repeating. O2 has, on paper, one of the strongest possible positions to offer SGP.32-managed IoT connectivity. It has a large installed IoT base, national network infrastructure, LTE-M, private-network experience, Telefonica's international footprint and platform, a large business channel through the Daisy combination, and wholesale relationships that could resell it at scale. Yet there is no publicly named O2 or Virgin Media O2 SGP.32 product for a buyer to point at. The capability is implied by everything around it. The proposition is not on the shelf.

For a deeper, term-by-term treatment of the standard itself, see the dedicated SGP.32 reference and the eUICC and eSIM standards library.

Accident or design?

There are two readings of O2's IoT invisibility, and the truth is almost certainly a mixture of both.

Invisible by accident

On this reading, the invisibility is a by-product of history and structure. Business units and brands have been reorganised repeatedly. Legacy web pages describe propositions launched years ago. Capability is split between O2, Virgin Media O2, Telefonica and partners, so there is no single obvious door. Private networks and broad digital transformation have taken over the marketing narrative that public IoT connectivity once occupied. And the migration to the new O2 Business brand is recent and probably incomplete, so IoT simply has not resurfaced yet.

Invisible by design

On this reading, the invisibility is rational. Huge infrastructure contracts such as smart metering need very little public marketing. Wholesale partners acquire and support the smaller customers that would otherwise be expensive to serve directly. O2 monetises network traffic without carrying the support overhead. Partners provide the hardware, portals and applications. Direct sales concentrates on large enterprise and public-sector projects where relationships, not web pages, win the work.

Both readings point at the same conclusion. O2 may be one of Britain's largest IoT connectivity providers without being one of Britain's best-known IoT providers. It supplies the network, the wholesale access and the major infrastructure, while aggregators, integrators and specialist IoT companies supply the visible service. That is not a company that missed IoT. It is a company that decided, or drifted into deciding, that being the layer underneath was good enough.

The most likely truth is that O2 did not sit down and choose invisibility as a strategy. It made a series of individually rational commercial decisions: to lean on wholesale, to chase prestige private-network projects, to fold IoT into broader enterprise sales, and to reorganise its business brand more than once. The cumulative effect of those decisions was a large IoT business with no clear public face. Invisibility by a thousand sensible choices is still invisibility, and it is reversible if the will is there.

What O2 could do next

The capability is there for O2 to become a far more visible IoT provider if it chooses. It has a large installed base, UK network infrastructure, Telefonica's global footprint, Smart Connect, wholesale relationships, a large business channel through Daisy, 5G Standalone, LTE-M and real private-network experience. That is an unusually complete set of ingredients.

Packaged well, it would support a clear SGP.32-managed IoT service, local O2 operational profiles, international Telefonica profiles, genuine multi-network resilience, partner-managed options, hardware and connectivity bundles, integration between public and private networks, and eventually satellite connectivity as standards-based non-terrestrial IoT matures. The question is not whether O2 can do this. It is whether O2 Business, in its new complexity-reduction guise, decides that IoT deserves to be a headline proposition rather than an ingredient in someone else's digital-transformation sale.

The Daisy channel is the wildcard

The most interesting variable is the Daisy combination. The merger that created O2 Business brought in a large channel serving small and medium-sized organisations as well as enterprises, along with billing, support and managed-service capabilities and existing relationships with tens of thousands of business customers. That is exactly the kind of machine that could take a packaged IoT connectivity proposition to the mid-market that O2 has historically pushed towards partners.

So the rebrand could cut either way. It could be the moment O2 finally makes IoT accessible to smaller buyers through a broad, service-led channel. Or IoT could be judged too specialist to earn a headline in a business built around simplifying technology, and remain buried a layer deeper than before. The new O2 Business website, as it stands, points towards the second outcome, but the migration is recent and unfinished. Which way it settles is one of the more consequential open questions in UK IoT connectivity, and it is worth watching over the coming quarters.

For now, the honest summary stands. O2's IoT operation is almost invisible to the market, yet its connections are growing faster than the parts of the mobile business everyone can see. That is a strange and interesting place for one of the country's largest IoT networks to be.

Frequently asked questions