It’s time to face telecom reality, people. 5G failed. 6G is going to do the same, for the same reasons. Telcos have moaned about declining revenue per bit for decades, with no resolution to their problem in sight. We hear the same sort of stories about new revenue opportunities, new ways of managing costs, that we’ve heard in the past, and all these end up where the past ones did. There’s a growing sense that the only salvation to telecom is some mechanism for subsidization. That may be true.
The question now, then, is what would have to happen in order for telecom to get on track. I’ve worked in the space for decades, worked on initiatives that were part of that past-failed-hope cascade and seen not only the failures but the causes. Here’s my personal view of both those causes and the potential remedies.
Problem number one; telcos have surrendered to vendors. In every single telecom initiative I’ve ever been involved with, and in every one that telcos themselves have talked about, one singular truth has emerged; telcos let vendors in, and let them take over. Vendors have outnumbered telcos, vendor spending on projects has overwhelmed telco spending. The rules of the telecom game are not set by the key players, the telcos, but by people who want to profit from them. Thus, the rules for what should be telco-centric initiatives are built to favor others.
There are two reasons for this, IMHO. First, vendors see all telecom projects as either a direct risk or a direct opportunity for them, and they play the projects with their own goal of promoting the latter and avoiding the former. Telcos do not see things that way, because vendors have always defined success in terms of increased sales, and telcos in terms of improved supply. Second, telcos have a history of regulatory restrictions on cooperation. I’ve seen them walk out of meetings because they felt there was a risk that the topic that was raised could risk their being accused of illegal collusion. Even where that risk doesn’t exist today (and it still does in many markets because the political winds blow regulatory policy), it’s hard to shake the institutional mindset.
Solution: Regulations have to free telcos to cooperate in their planning, and telcos have to frame the structure of projects, industry groups, standards groups, to ensure they have staffed the teams involved to dominate the procedures. Current activities cannot be saved; new ones will be needed.
Problem two, supply-side bias. Telcos are used to building into a well-understood opportunity, a need that only has to be fulfilled to succeed. That collides with the simple fact that connectivity is a commodity that can be differentiated only by price. Think about all the service features proposed in the past, from turbo buttons to token pricing in AI. All of them are ways to get more money for connectivity, and none of them were adopted because nobody wanted to pay it. If you want to do supply-side planning, you need to focus on an area where pent-up demand exists, and where pricing power can be achieved. Connectivity isn’t that place.
Solution: It’s apparently clear to almost everyone but telcos that the only place where new money can be obtained is in the area of real-time services. The problem, for those few telcos that actually see this, is that they fall into thinking of it as a connectivity problem. Build low-latency features into 5G or 6G and you’ve created new services. The problem is that real-time isn’t a connectivity problem, it’s an application problem. We live in a world where you can’t make money connecting people, you have to make it by connecting people to applications, and applications connecting to other applications. Absent the applications, you have nothing. Telcos need to focus project work on application-building. Example: AI connectivity is a stupid effort, because it’s not AI that’s important, but AI applications. Applications are what build willingness to pay.
Problem number three, failure to realize telcos’ advantage. You can’t be effective in the market if you can’t play to your own strengths. Telcos have two. First, they’re free-cash-flow machines, meaning that from a stock perspective they’re usually dividend stocks. With Wall Street increasingly worried about tech valuations in general and AI in particular, that’s a nice asset. You don’t have immediate pressure to grow revenue as long as you can credibly sustain dividends.
Second, telcos have a very low internal rate of return (IRR), which means that they can make a business case for projects whose ROI is relatively low. That means that they can tolerate, and even gain from, investments that other companies would find highly unattractive.
Solution: Telcos need better investor-relations messaging to accent these points, but they also need to realize that they are in a position to invest in things critical to the future, which means that real-time stuff I mentioned earlier. Forget the notion of “AI-RAN” where GPUs sit in towers. The real key location asset is the metro, and they also need to refresh what “metro” means. Originally, a metro area was a combination of a population analysis tool and the basis for dividing local and long-haul services. Today, what we should think of as a “metro” is a place where content delivery and Internet peering connect.
The key to real-time success is to both achieve economy of scale and latency control at the same time, for your hosting points, and then add value to hosting by combining pure hardware with platform software and APIs that can drive development. To do this, telcos need to rethink their whole approach to APIs, shifting from today’s “how-do-I-sell-connectivity-differently” mindset to “how-do-I-promote-applications”. Hosting alone is a commodity business, and while the telco IRR makes that OK for a time, telcos need to escape commoditization to truly stabilize their business.
So, you ask, could telcos do this? I’ve worked with a lot of them, and I think they could. But I agree with those who would point out that, as I’ve said myself, they could have done this before and didn’t. Why now? There is no better answer for this question today than there was twenty or thirty years ago, other than perhaps a greater degree of urgency. And, sadly, most telcos I’ve chatted with don’t think they could do this, which is likely the greatest barrier of all.
