There’s no shortage of material offering advice to communications service providers (CSPs), often called “network operators” or “telcos”. I’ve contributed my share in these blogs, in fact. One organization that’s specialized in the CSP space for decades, the TMF, has recently worked with Tech Mahindra to produce a paper called “Reigniting telecoms growth: A playbook for CEOs”. It probably won’t surprise you to hear that I don’t agree with everything it says, that I have a slightly different interpretation of some of its points, and that some of it doesn’t match what I hear from the telcos themselves, but it’s worth exploring both the paper and these differences.
Let’s start with the positive. The paper implies a new structure for CSPs, one that divides “services” and “infrastructure” as “ServiceCo” and “NetCo”. Many telcos I’ve talked with say that’s what they’ve backed into over time, but their real organizational boundaries still conform to the CIO/COO/CTO divisions of the past. In fact, over three-quarters of telcos say that their current organizational structure is “an impediment” to growth, and almost all say it’s obsolete and sub-optimal. I think this new model is one of the best ideas the paper presents, but I also think it would be enormously difficult to bring about. We’ll see how successful the paper is at framing an approach to the new model.
The next point the paper makes is the ever-popular (with telcos) theme of disintermediation. It points out, correctly, that telcos have failed to address the evolution of network services from connectivity delivery to experience delivery. Where I have a problem with this piece is that it’s far from clear that the problem could have been prevented. In the US, for example, the Telecom Act effectively separated the old and new models into separate operating units, and there was always a risk that investing in experience services would end up generating assets that telcos would then have to wholesale to competitors. The fact is that regulators didn’t understand what the whole privatization movement created, and didn’t structure regulations to suit the future.
Now, says the document, the “playing field is changing again”, and the focus of the change is B2B services. A bar chart shows that telco estimates for revenue growth divide neatly based on whether the telco believes B2B or B2C will dominate. Those with modest revenue growth expectations feel B2C will dominate, and those who think revenue growth will be more robust are universally depending on B2B for that. There’s an element of reasonability here in that enterprises tend to value CSP/telcos more, but this is the first area where I find myself, and the telcos, disagreeing with the document.
First, the divisions of B2B and B2C are both business-centric, and unless you decide that the latter includes everything in the content delivery, gaming, and social media spaces the two categories miss the big drivers of change in networking as a business. Second, the adoption of cloud front-end technology for business applications has actually shifted enterprises more to what looks like a B2C model. This trend is now spreading to the use of the cloud/Internet to support workers, even in branch locations. You could easily argue, as 80% of telcos tell me they do, that “business services” in the sense of things like enterprise VPN services, are going away over time.
According to the document, CSPs think their best opportunities for revenue growth lie with connectivity services (the stuff that’s already in a slide in terms of profitability) and “end-to-end” services, which seems to imply a role as an integrator of connectivity services. This fits pretty well with what the telcos say at the senior management level, but at the planning level (down a couple of management layers) telcos believe that neither of these approaches will succeed. I don’t think they will either.
The strategies that would align CSPs with opportunities are a bit difficult to extract, but I think the report is recommending a different kind of layering, with the “NetCo” at the bottom, and the “ServCo” (they do stick in “InfraCo” as another company layer in a later figure) making up three layers above. The layers are no more than interesting because, IMHO, they don’t make the obvious tie with AT&T’s “facilitating services” notion, and that’s the heart of the problem I have with this part of the report. The report builds on the layers with the idea that telcos will need to respond to market shifts “by selecting positions that set them up for scaled partnerships and/or horizontal leverage”. Aside from the fact that this is hardly real-person-speak, it doesn’t offer a real answer to the question “Why partner with a telco?”
The next page provides the unreal answer, which is that ServCo’s service is connectivity-as-a-service. Gang, that is simply nonsense, a reprise of a failed notion that has a decade or more of negative history attached to it. They do toss in multi-access edge computing (MEC) as a kind of an aside, but there’s nothing about how to get there. Instead, the paper shifts to what could fairly be called “pap for the management masses”, meaning the usual organizational and management practice recommendations that have no chance of success without a link to a practical technical path forward. And, in fact, a couple pages later the paper admits that the telco model has to accommodate “value shifts to cloud services”, which seems to admit that their partnership strategy is to accept the cloud providers as OTT disintermediators of telcos.
There is a technical link in a discussion of the role of APIs, but the problem with an API approach is that unless the assets you expose are highly valuable to the higher-level partners, what you’re doing is facilitating the conversion of the only asset you have, connectivity as a service, to a wholesale piece of some higher-level element. It’s hard to see how that doesn’t make disintermediation worse.
The rest of the sections, from “Strategy” to “Tools” can be summarized as a description of those management platitudes again. The problem is that you can’t worry too much about managing a process whose technical steps you don’t support and probably don’t even understand. If the value is shifting to cloud services, and given that telcos have zero success in being cloud service providers, why worry about how you’re going to manage the cloud-based transformation?
The positive side of the paper is that the point on a restructuring of the telco organizational map is surely a good idea. Companies tend to do what they’re organized to do, and organizations are often barriers to change because changes demand a different planning and approval flow, as well as different kinds of collaboration among workers. However, the organization of telcos was set by the services they expected to offer. To expect changes in organization to change service offerings automatically does not follow. The telco network defines what services it can expose, and before you decide on optimizing organizations and practices, you need to plan and frame infrastructure.
That’s the negative side. In the end, layers and APIs and the other elements of the paper’s technology plan (sparse as it is), are nothing unless there’s something different to offer. MEC is the only thing that’s cited as a plausible difference, but how can telcos play in that space? “Carrier cloud” is no longer accepted as a goal by operators; only 8% of telco planners say they believe it’s in their future. So do they partner? What could the telcos bring to the table to ensure they got reasonable margins, and why would buyers go to them for cloud services that buyers know is really supplied by a cloud provider under a wholesale deal? Instead of MVNOs, the paper seems to want operators to become CVNOs, cloud virtual network operators.
Why did the paper end up as it did? I think the answer lies in the constituency it’s aimed at. The TMF is a conservative-telco body, populated largely from the OSS/BSS side of the business where the Field of Dreams is still going strong. They want to hear that the future of telcos is still the connection services of the past, perhaps exposed via APIs and tweaked a bit, but “connection as a service” is still connection. There’s lip service paid to edge computing and APIs, but only insofar as they define infrastructure evolution driven by…you guessed it…connection services. That sort of future has already been discredited in the present, and it’s sad that the paper didn’t acknowledge some of the recent notions (like facilitating services) that telcos themselves have already started to consider. The paper tells their audience what they want to hear, and that’s a natural thing to do. But it’s not helpful, and with just a bit of re-tuning, it could have been. What that “could-have-been” might look like is a topic I’ll cover later this week.
