VMO2’s cost-cutting, Schrödinger’s ‘Starlink’

Okay, what have we got – worth a comment? Let’s try to keep this short-ish. So there’s serious cost-cutting at VMO2 in the UK – £800m in annual operating costs by 2029, including £400m as soon as next year, and £300m from workforce reductions. It comes on the back of a grim quarter (revenue down 5.7%; EBITDA down 2.6%; £197.3m out-of-pocket cash-flow; fewer broadband and mobile connections), which contrasts with the broader tier-one telco tale that has come out of the quarter. It’s put a halt to its full-fiber build, which again bucks the trend (certainly in the US), and follows its apoplexy at CMA decisions in the UK about BT’s cut-price TalkTalk rescue, and its own “logical” £2bn Nexfibre joint-venture takeover of alt-net Netomnia. Its 5G infrastructure investments are safe, it seems. Its strategy is basically: less operational duplication, slimmed-down portfolios, better business synergies, smarter focus, and probably harder work; plus AI holds real promise, it says, to automate service and network functions. “This feels like a last roll of the dice to rebuild confidence,” said telco analyst Paolo Pescatore, always good.

As well, there’s hard news and hot takes in the satellite space. Of course there is. Except while SpaceX pours rocket fuel everywhere, to light rumor and bluff, and kicks tyres and “scares” telcos into granting some “backdoor” entry, the old guard is going about its business. Sort of. AT&T and AST SpaceMobile have demoed handover between terrestrial and satellite networks. Which is different, it seems, to what current D2D services (including Starlinks’) support: where they switch between, when terrestrial drops, but don’t necessarily preserve the session. In AT&T’s case, it provided the ground system, AST provided the satellite set-up, and Ericsson and Nokia provided the radio gadgetry on either side, respectively. Probably, AT&T (and AST, and the others) needed the good-news NTN story; there’s certainly not a lot in the release – about the type of connection, the time of handover (the ms figure is only really a description of what Ericsson’s tech is designed to achieve), any mention of interruption or degradation, whether it worked in reverse. It’s proof of hybrid mobility, then, not of seamless commercial service.

As well, some more reaction to yesterday’s rumours about SpaceX buying Boost Mobile, as cribbed from social media – in case you’re interested. Dimitris Mavrakis at ABI Research said: “Rumors about SpaceX building its own mobile network continue to get… confusing… Boost Mobile is an MVNO running on AT&T. All three US carriers have refused an [MVNO] partnership – [so] SpaceX will need to build its own network, [which is] expensive and time-consuming. But buying an MVNO without its own assets – which is also running on AT&T? That makes no sense.” Richard Haas at LS Telecom UK said: “It’s hard to take this seriously. A few weeks ago, the rumor was SpaceX was in talks to buy capacity on 80,000 towers… Musk has [also] alluded to mysterious mobile plans, saying it would ‘develop entirely new pieces that fundamentally alter the strategic landscape’ – implying cellular industry veterans wouldn’t understand the company’s approach. What’s really going on? Who knows?” At this point, we’re dealing with Schrödinger’s Starlink (fat cat?) – simultaneously a fully fledged operator, an MVNO, and everything between.”

The last point captures some of the confusion, whether offence or bemusement, at least in telecoms – that Musk thinks he can rewrite the physics, or else just buy up the industry. Meanwhile, there are enough fan-boys out there, likely with irons in the AI bonfire, that think he’s god. Just look at the way he’s dressed up like a chiselled A-lister on their AI posts. Groan; end rant.