Nearly 100 different firms are rumoured to have looked over Nokia’s books to see what its ECE division was really putting away – about £170m per year, as it turns out, growing at a double-digit annual rate for a decade at least. So it is funny for Nokia, in a way, if a cut-throat VC/PE profit merchant (per the sector’s reputation) thinks it can make a business out of it. Clearly, Aurelius thinks it can (and would not describe itself in such terms). ECE is to be placed into its ‘opportunities’ fund when the deal closes in early 2027, with a mission to expand at the ‘campus’-end of the private 5G market. The London/Munich-based outfit paints itself as a specialist in “complex, global carve-outs”. Very clearly, this will be just that: 1,000 staff in 60-odd countries – is unlike most niche carve-outs, even from big firms.

All told, this looks like a good deal for all parties. Following its FWA sale to Inseego, Nokia has now got rid of at least two of its basket-case ‘portfolio businesses’, potentially – and pretty sharpish, too. Its microwave and ‘site implementation’ units remain up for grabs, it seems; but chief exec Justin Hotard looks like a “guy that can execute”, and the Finnish firm’s focus is simplified around its two large-scale ‘supercycle’ operations: ‘mobile infrastructure’ (MI, selling cellular to telcos) and network infrastructure (NI, selling fiber to anyone who wants it, including desperate and deep-pocketed AI and cloud companies). Meanwhile, its ECE division – which was bold enough to open up the industrial comms market for the whole telco industry, and unafraid to ruffle feathers along the way – gets a new start.

And Aurelius gets its whole tech portfolio, we hear – and a way into the tech industry. Some details are unconfirmed: the DAC and MXIE lines are included, along with its industrial devices and applications. Its MPW system, for larger-scale core deployments, is not referenced in the press material, but is understood to be included in the deal as well. There is no mention of its parallel drones business (Nokia Drone Networks) anywhere, which grew out of the ECE stable, and occupies the same industrial headspace, but is as often associated with public safety networks – notably with deployments by Citymesh in Belgium and Swisscom in Switzerland. It is unclear as well which products will be cross-licensed. MPW, say, mixes core elements from both its MI and ECE units.

The expectation is that ECE will continue to sell Nokia radios with its campus sales under its new parent. Relating to Aurelius’ ‘carve-out’ strategy, there are interesting portfolio studies on its website: Dutch IT services company Getronics saw its headcount grow from 2,700 to 4,700 over five years, until Aurelius’ exit in 2017. It has a long list of (live) interests in tech-related distribution channels – as written in the coverage linked below. Which gets us into a conversation about the state of the market. Nokia is selling ECE because private networks are a “systems integration business” – rather than a box-shifting discipline to a limited number of customers. The irony is 5G, as it was specified and defined, is an enterprise tech, to all intents and purposes, for different enterprises in different industries.

But it needs to be made custom each time; the old 80/20 rule is traditionally applied for generic/bespoke installs. At the time, 12-18 months ago, Nokia’s decision to sell its ECE unit seemed shocking because it had worked so hard, and made such progress – and dared even, in earlier days, to go against its old operator channels, dragging their feet, to sell directly to enterprises. Ericsson, it might be noted, was way more cautious about upsetting its old compadres. How the tables have turned. The ECE unit has flip-flopped on its channel strategy, and settled lately on a rough 30/70 division of labour, between its direct/indirect channels – with the former taking closer charge of larger-scale deployments. These stretch to 50-plus sites in the largest cases.

But that – use-case driven integration, direct or indirect – does not fit the telco model. That is the only conclusion; you can’t teach an old telco dog new tricks – especially when there is a simpler and richer opportunity at hand. It does not work for Nokia, both-footed in the AI cloud space, and, per the Vodafone discussion last week, it does not work very well for operators either – now they have public network slices to sell. Not every operator thinks like this, but Vodafone’s presentation at its investor day last week looked telling: lots about slicing, hardly a mention of private networks. They will revert to type, and it makes sense in ways. And campus-based industrial enterprises, which don’t want slices and AI-RAN for OT workloads, will also revert – to the types of channel partners ECE has stacked up, which Aurelius knows well.