EU telcos want Big Tech subsidies. Telcos worldwide are saying that their profit per bit is threatening to go negative. Vendors who focus on the industry are reporting misses in their numbers and are revising their outlooks in a negative direction. Does this sound healthy to you? I’ve talked about the issues behind this for years, but of course I don’t get to make industry decisions, only recommend things and comment on results. What do telcos themselves think about the best path to solving their economic problems?
This is obviously a difficult question to address, because there’s a lot of concern about making negative comments or even admitting that there are issues to address. I’ve worked with telcos for decades, and surveyed them for that same period, and I’ve always promised total confidentiality. I’m hoping that continues to get me honest answers, and I’ve also tried to analyze what I hear from multiple sides to prevent bias, but you need to know this is sensitive stuff and there’s really no way to know how much people will disclose. Reader, beware.
Is there a problem? Of 47 telcos I’ve managed to get comments from in the last three months, all agree that there is one, and all agree that the problem is that the traditional business model of operators has failed them in the age of consumer networking. Today, content delivery and B2C activity, driven by consumer broadband Internet, has gutted almost everything about that legacy model. Telcos say bill and keep, the foundation of Internet economics, has encouraged ISPs to focus on the producer side of everything, which requires relatively few very fast connections to source providers, and leave the consumer game, with millions of small users who are highly price-conscious, to the telcos. There is competition between telcos and cable companies that keeps margins on consumer broadband thin, so there’s no revenue-side escape.
This basic “truth” (which is true as far as the telcos are concerned, which is true enough if you’re analyzing what they’re likely/willing to do) means that new revenue avenues are suspect. Staying with that 47 telcos, 45 believe that the business model of networking today doesn’t favor them in the introduction of new services. They think that anything that is proposed is going to have a small number of super-buyers that will generate boatloads of revenue, a mid-range set that has similar revenue potential but a larger number of users to support in order to get it, and the masses that are unlikely to be more than marginal. They see the first group cannibalized by opportunistic specialists, the cloud providers grabbing most of the middle group, and their own companies left with the rest…just like broadband.
This is why telcos are telling me that my theory of service opportunities in areas like edge computing and IoT aren’t attractive to them. It’s not that they don’t believe there is opportunity there, or that they don’t think they could address it. It’s that they’re concerned that they will be mandated to serve the broad market when the broad market will never be very profitable.
The telcos who believe in Big Tech subsidies (14 of the group) say that Big Tech was essentially a creation of the same business-model shift that compromised telcos’ own financial situation. The old model, with settlement among participants in a service, included implicit subsidies in that settlement process. The new model abandoned implicit subsidies and so requires the subsidies be made explicit.
OK, then, the view is that a revenue-based solution to the telco business model challenge isn’t viable. How about a cost-based solution? If profit-per-bit is the problem, producing bits cheaper could be a solution, right? Yes, right, but we have to be a bit more sophisticated than saying “producing bits cheaper”. Carrying bits is the operators’ mission; Big Tech is the one producing them. Carrying bits is part transport, which operators are pretty efficient in handling, and part access, which is by far the most costly piece of the puzzle. Access, say the 47, is really about something I’ve often blogged about but something rarely talked about elsewhere—pass cost.
If I want consumer broadband, my operator isn’t going to start trenching from their hub location to my home; the cost would be prohibitive (and a lot of people along the way would get pretty upset). The reality is that access starts with a “positioning build-out”, where access infrastructure is laid out like a circulatory system with major and minor arteries. Those minor ones “pass” prospective customers, and from each of them it’s possible to reach users with a relatively small stub (“capillary”) connection. If you don’t have your passes lined up, marketing services is going to be not only a waste but a generator of angst. Nobody wants to hear about a super new service that they can’t get in their own area for a year, and they get mad at the player who tells them.
Technology improvements have made it possible to offer consumers broadband connections for less than $100 per month, when in the 1990s half the average current capacity would have cost enterprises tens of thousands of dollars per month. Sadly, the ROI on that mass broadband is lower than it was on the few thousand connections we might have seen in the 1990s. Perhaps even more sadly, access infrastructure has some baseline requirements that aren’t easily supported using improved technology. Whether you’re running twisted pair or fiber optics, you have to pull cable or dig trenches, make connections, and so forth. As we advance access technology we pick off all the places were basic technology improvements cut costs. Other areas are likely to see cost increases. Bet you can’t dig a trench in 2023 for the same price you could in 1995. Thus, if it’s hard to make money today on broadband, it’s going to be harder in the future to even control costs much less reduce them.
What does this have to do with IoT or edge computing? The answer is that you have to be able to market/sell those things too, and to offer means to be prepared to deliver. I can sell or I can market something. If I do the former, I can focus on the very specific places/customers I’m actually prepared to serve, but my cost per prospect is high at best, and if I have to build awareness and interest, it could be very high. I can market, meaning use mass media to reach prospects, but if I do that then I lose my ability to ensure that those who I capture can actually be served given my current state of deployment. In effect, I have to “pass” my market area in order to start marketing, and that’s costly.
A few telcos, just a bit over ten percent, think that the answer to all of this is the greater use of optical routing and DWDM in creating a low-latency, high-capacity, IP VPN infrastructure. This could encourage enterprises to extend their current IoT applications, which host control-loop applications on site near the facilities, across multiple facilities. That, in turn, could develop a market for edge computing that the telcos could exploit because it’s their enhanced VPN service that created it in the first place.
There’s potential here, but of course it’s not a slam dunk. Telcos are still amateurs when it comes to hosting services, and even more so when defining a set of platform elements to support edge computing would be required. Would creating an edge opportunity be difficult? Yes, but at least possible, and that’s progress for the telco world, which sorely needs it.
