Light Reading often has interesting stories, and one that fills that mold is on Ericsson, a company dating back to the year George Armstrong Custer fell at the Little Bighorn. Says the tagline of the piece, “As it celebrates its 150th anniversary, Ericsson must contend with an ongoing slump in customer spending, 6G and geopolitical uncertainty, and the relentless march of AI.” Yes, that’s true, but hearken back to some points I made in yesterday’s blog. In the telco world, how many “must contends” have we managed not to contend with?
The first headlined section of the piece is “Toward a hyperconnected AI world”, where “later-stage 5G and pre-6G technologies are planted in smart glasses, robots, industrial machinery and other objects.” The problem with this is that it’s not at all clear that there’s a business case for any of those things. Even if there is, there are a lot of pieces needed to build the justifying applications, and Ericsson doesn’t even make them all. AI is, in the picture, not the driver of change that will inevitably succeed, but another of the costs we’d need to justify in our business cases.
There are a lot of things you could do with AI, but there were a lot of things we could do with ISDN, frame relay, ATM, 5G and so forth. Utility isn’t the same as suitability-for-purpose, and that in turn isn’t the same as creating-justification-value. Should Ericsson assume that somehow the demand for all that wonderful stuff, the “tailored user experiences” the article quotes, will actually develop? Remember that the same piece, a couple paragraphs later, admits that it lost ground to Huawei in the past. Huawei is a price leader, and you generally can’t focus R&D on being price-competitive while you’re making speculative bets on service evolution. Particularly in an industry where past bets have gone awry more often than they’ve paid off.
5G is really the issue that everything for Ericsson, and for the industry, revolves around. Any mobile standard has to balance the “needs” against the returns on fulfilling them. 4G networks, the piece notes, needed more capacity to cope with consumer broadband and smartphones. Yes, perhaps, but how does that pay back in terms of service revenues? Telcos started their disintermediation complaints less than a decade into this century, and showed me their now-classic revenue-per-bit charts plummeting in 2012, five years before current CEO Börje Ekholm took over. Where was the 5G contribution to fixing that? Obviously, it didn’t work, so why would Ericsson (and Elkholm) do better now?
The problem that Ericsson has, that telcos have, that the whole telecom industry has, is that “The hope was that 5G would be used outside smartphones to power self-driving cars or allow robots to perform surgery. So far, it has not happened at scale.” True, but the fault lies in the “hope” part. Things don’t happen in a global market because someone hopes they will, they happen because someone drove them to happen.
Why would businesses, people, spend more on telecom services or anything? Remember, as the tagline said, we’re in a period of geopolitical uncertainty. They’d have to have a good reason, which is another way of saying that they’d have to make a business case. How long would it take for self-driving cars to replace vehicles now on the road? How long would it take before we’re letting robots clean our living spaces or take out our appendix? Candidly, anyone who believed that any real progress could be made in any of those areas in a few years was dreaming, not doing market research. And, candidly, what operator would make a massive investment in infrastructure today in expectation of a return more than three years off, at best?
Why did PCs succeed? They were revolutionary technologies and the uptake of PCs by businesses was fast enough to cause IT spending to grow far faster than GDP. The answer is that a PC was a more personal way of doing stuff we did impersonally, and also a very cost-effective way. Smartphones were based on the assumption that people would want the Internet experience they got at their desks when they weren’t at their desks at all, or at any desk. We had opportunity, demand, in place. 5G robotic surgery? Self-driving cars? We have some robotic surgery today, some autonomous vehicles, but far from enough to create pent-up demand, and in any event, can you really run a vehicle in traffic through a mobile network connection? Can you do surgery over such a connection, given that it might fail at a critical point? There are some cases where both these might be essential, but enough to justify mobile infrastructure spending on a large scale?
Ericsson’s problem is everyone in tech’s problem. They understand technology, but not why someone would need it. They “hoped” that someone would develop the applications that would justify the need, when that unnamed force would have to make their own business case for their work and products, and likely they’d have others who’d have to make their own business cases first. What is enterprise tech spending doing these days? Modernization and cost management, because there are few tech projects that can make new business cases, and thus contribute benefits to offset their adoption costs.
Why is this such a problem? Some enterprises tell me what they think they could use to build new business cases, but they can’t focus vendors on these long-term requirements. Some say they can’t develop use cases for new technologies themselves, they’d need vendors to take the lead, and most don’t do that. It goes back, I think, to the short-term-sales focus. A number of my friends in sales tell me that they need to make quotas for the current year, the current quarter. Pushing ideas that would take years of work to turn into a sale just isn’t in the cards. People do what they’re paid to do, after all.
And this stuff is new, hard, and risky. But it’s also the only way, in the long run, to make the pie bigger. Ericsson, like most tech/networking companies has been cutting jobs to preserve profits. Smart for them, essential in fact. But why then isn’t it also smart for their prospects and customers? How did Ericsson sustain hopes that the strategy that they’ve themselves adopted to protect their business would be ignored by those they were trying to sell to? The same way other tech companies, and AI companies, do today.
Some enterprises tell me that they think “tech consumerism” is a contributor to this problem. More and more technology is sold to individuals these days; fifty years ago, people didn’t own computers and smartphones didn’t exist. Is it possible that tech, increasingly dominated by consumers, has started to market in a consumeristic way even to businesses? Forget actual business cases and focus on “everyone else is doing it”? If so, it might explain how vendors forget to consider the business case of their buyers, the complexity of modern tech value chains.
Ericsson is selling to a vertical that has notoriously long periods of depreciation, which means that it takes a long time for a technology to become eligible for modernization. If its buyers are cautious about proving the value of technology change, they’re doing the same sort of profit optimization that Ericsson itself is doing, and Ericsson simply has to contend with that. So do all vendors, in the long run.
