There are a lot of reasons why a market sector gets into trouble, but nearly all of them stem from a single factor. When buyers can’t make any new business cases for a technology, that obviously saps opportunity to increase sales of that technology. Vendors then pin their hopes on the “refresh cycle”, the regular replacement of aging gear, for their sales upside, but they forget something. Their buyers are under the same pressure to increase profits as they are, and so they want cheapness when they shop for replacement gear. And, nothing is cheaper than what you already have. Light Reading cites an Analysis Mason report and other sources in their story on “skinflint customers”, but it may not go far enough in its warnings.
I have regular input from 88 operators, and all 88 of them tell me that they are stretching out their capital purchases by waiting until old gear actually generates problems before replacement. All 88 also tell me that what they want in the replacement gear is lower cost. The upgrade to networks created by 5G has largely played out, and it’s time, say the operators, to tighten the belts.
Until? The popular answer (from 54 operators) is “6G”, but the remainder are saying that their hope is that something better comes along. “Better” here means something actually offering ROI. 5G was a loss-leader evolution, says this group, and you can’t recover from that by transitioning into another loss-leader evolution. “Loss leader” is supposed to lead to something more profitable,” an operator planner told me. And 25 operators said they were thinking of sitting 6G out unless it led to a transformation on the RAN side that pushed a refresh to the handset space, and they’ll push standards to avoid it creating an artificial driver for a refresh of RAN infrastructure. They did not believe there was any ROI coming with 6G, and the Light Reading story mirrors this.
What operators want, though, may be the impossible dream, a software-feature explosion that drives new service revenues. What makes that especially difficult to achieve is operator rejection of non-connection services. “My management wants new services that end up looking like connections,” one planner complained. “You can’t propose anything that an OTT could do too.”
Another Analysis Mason report, cited this time by SDXCentral says that telcos are in a unique position to exploit what it calls “multicloud NaaS”, which is the use of APIs by enterprises to compose their own IP network services. The API concept, at least where what the APIs expose is the features of connection services, is rejected by 31 of the 88 operators. The same planner I quoted above offered a reason: “So an enterprise can compose IP connectivity. What does that give them that they don’t already have? What do they compose that’s different? Why would they pay more to get the flexibility to do what they already do?”
Who is responsible for creating the demand for the new features operators want? The buyer, say 66 of the 88, which reflects, I guess, the field of dreams mindset. Only five thought operators themselves were responsible; the remainder said vendors had the responsibility. Even in the question of the specifics of the APIs, operators pointed to others to take the steps. I would argue that the concept of “multicloud NaaS” is a creation aimed at framing a “requirement” that operators (who have to interconnect for regulatory reasons) could fulfill instead or cloud providers (who have no such responsibility). The world has to frame its future in connection terms and feed it to the telcos. Ha!
One thing I think is clear, and that is that the future of network service evolution has left the telco central office building. It now resides in data centers that seem increasingly unlikely to be owned by telcos. But, as I pointed out in my blog yesterday, the cloud providers don’t seem to be interested in driving feature evolution of services either. Who then does that? The NFV ISG, a mixed group of operators and vendors, has failed to create even the framework for future hosted features. I’ve participated in a number of standards initiatives aimed at feature transformation, and they’ve all stalled on the fact that everyone involved wanted to avoid discomfort or cause others to be uncomfortable. Avoidance is an inhibitor, not a driver.
Could there even be a service or feature set that could drive telco revenues upward, justify new equipment spending, and create industry-wide happiness? I think there could, but the problem is that the service would have to be something that altered the way we work and live so fundamentally that it would change what “connection” means. The sheer scope of change means it would have to involve a lot of parties, and this in a space that, for decades, has been unable to get any movement even within a single party.
“All day I faced the barren waste, without a taste of water,” so the song goes. When something is scarce, all you need to generate demand is to offer it. Today, we have to make water smart, flavorful, fizzy, sexy, or whatever. In the dawn period of telecom, connectivity was scarce and so offering it was all that was needed. We’re in the designer-water phase of connectivity, but telcos are still offering a bucket and a dipper to the imagined thirsty, when we have connectivity nearly everywhere. I had the experience of standing at the foot of the mast where Amundsen moored his airship before a flight over the North Pole, where I had to be guarded with a rifle from the polar bear danger, and I had Internet access. Simple connection doesn’t cut it, APIs or not.
The LR article is right. We are facing a long capital starvation in the telecom sector. The SDXCentral article is right. We need to compose our technology world, which includes connecting the pieces. We need to have the pieces, the vision of that technology world, to have the first of these rightness things give way to the second.
