Ever notice how many maxims we regularly hear come from sources almost nobody recognizes? Often, in a slightly different form. One is “this too shall pass away”; the actual poem by Theodore Tilton shows the original form is “Even this shall pass away”, but whatever form you like, there’s growing sentiment that telecom, and maybe networking in general, has seen its best days. Is it justified?
The Light Reading piece I quoted above isn’t the first indication of this, of course. We’ve seen a lot of recent announcements of this vendor or that exiting the telecom space. The immediate problem is that the necessary part of the 5G standards adoption cycle (the RAN) has largely passed, and what’s left isn’t going to warm many vendor hearts. Some hope that 6G will pump air and money into another build-out cycle, but there’s a longer-term problem that isn’t limited to telecom, or even to networking. The problem is spending, and on what justifies it.
There are two engines of telecom, and indeed of tech, spending, the consumer and business. For each, the engine of revenue is divided into the number of users (the total addressable market, or TAM) and the average revenue per user (ARPU). As the Light Reading piece points out, most telcos have seen significant price pressure, driving down ARPU, first in wireline services and now in wireless. Why? Because broadening the consumption of something, increasing TAM, almost always means lowering prices, because willingness to pay is limited for almost every consumer of anything. At some point, though, further price reductions don’t generate as much from TAM as they lose in ARPU.
The consumer market is driven more by sentiment than by formal justification. People buy things because it makes them happy, or at least happier. Businesses demand a formal analysis of spending, one that expects a return on any investment. While this formalism is more apparent in capital spending, the problem that telecom services have is that pressure on project costs is general, meaning that a new project has to show that the benefits exceed not only the amortized capital cost, but any incremental expenses the project generates. So, in the end, businesses will likely increase spending only when there’s a new benefit to contribute to the business case. They buy because it’s profitable.
Back in 1990, business services made up over half of global telecom revenue, despite the fact that there were way more consumers than businesses. Because access revenues (local exchange) were separated from long-haul revenue then, we know that businesses accounted for more than two-thirds of the access spending, and access is the most costly piece of network infrastructure spending overall. What changed this was the global push for “privitization” or deregulation in the mid-90s, which in the US gave rise to the Telecom Act of 1996 and the breakup of the Bell System. Using the US market as an example (the market I live in and have the most data on) what this did was to transform telecom from a single-entity regulated monopoly to a competitive jungle with (then) seven or so and (today, after the inevitable consolidation) three major players. Competition lowers prices, and because of competitive overbuild costs, lowers profits overall.
The greatest competitive impact was on business services, for several obvious reasons. First, they had more money. Second, finding and selling to them was easier. You can send a whole sales team to call on a Fortune 500 company, but you’d never be able to even phone up an equivalent-spending number of residential customers. Consumers, though, were lured to competitive toll carriers by price. By 2000, in the US, AT&T had lost half its long-haul revenue, and of course that’s all they then had.
The Internet, specifically the web, was the big disruptor then. Business data services were based on high-quality-and-cost digital trunks, DDS at 56 kbps up to T3 at 45 Mbps, and none of this was remotely suitable for consumer price points. DSL and cable offered much better speeds, and shortly fiber optics shifted from being a deep-trunk technology to access aggregation, and finally subscriber connectivity. As this happened, the price pressure of consumerism dropped price per bit precipitously, and even in 1990 it was obvious that something new—cellular mobile—would do better. Mobile customers exploded from single-digit millions to half a billion in the US from 1990 to 2020, but mobile ARPU dropped off over 40% in the 1990s, then another 20% by 2020. Through this whole period, both wireline and wireless business services were increasingly impacted by consumerism; business would not pay a giant premium for what was essentially the same service or (in the case of wireline) sometimes even a lower-capacity connection.
Today, this combative duality of business and consumer services is driving the bus. Telecoms have tried, in cooperation with their vendors, to find new services that businesses would pay for, but so far the trend has been to design business applications to take advantage of the Internet’s consumeristic service set. If you look at 5G and 6G in this light, you can see that the focus has been to provide QoS that the Internet’s best-efforts model could not, even as businesses worked to make best-efforts good enough.
Today, in the US, we have two broad-based telecom providers, AT&T and Verizon. Both companies have seen their business services decline to less than 20% of their revenues. Wireless accounts for over half of AT&T’s revenues, and almost three-quarters of Verizon’s, but consumers won’t pay more for wireless services, and in fact want to spend less. Only business services offer any potential revenue upside, and that’s limited unless you can identify new benefits to justify new projects that consume new, different, more costly, services. More on this tomorrow!
Vendors, as the article says, are fleeing telecom, because telecom isn’t profitable enough to sustain them, but is telecom’s limitation really so bad? What do you flee into that, in the long term, is any better? All of tech has suffered from new-benefit starvation for decades, as virtually every enterprise tells me. If there’s no new revenue source, you spend only where you can consolidate total costs more than you’re spending. That’s never going to produce a dynamic market.
