Nokia, having completed its consolidation of the NSN partnership and sold off handsets to Microsoft, now faces the big question—“Where now?” The “New Nokia” seems to be a network equipment company, but it seems to me that if you look at Ericsson’s quarter you’re left with the realization that they might not have enough network equipment.
Ericsson and Huawei are the two giants in the telecom industry, and the two companies are quite different apart from their national roots. Ericsson has been making a highly publicized move toward professional services, and Huawei has always been a price leader in hardware. Sure both companies have both capabilities, but what they ballyhoo is where they see themselves heading. But Ericsson’s quarter was weak, and when it was announced I speculated that their problem might be too much service and not enough product. If that’s true, it might be demonstrating what Nokia needs to do now.
Look out over the telecom landscape these days and you something that (botanically speaking) looks much like a long-neglected garden. For almost a decade now, operators have been telling the world that the Internet pricing model and the growth of consumer service based on all-you-can-eat was devastating their revenue per bit. Vendors listened, sighed in compassion, whipped out their order books, and said “Gee that’s a shame. How many boxes can we put you down for this quarter?” Cisco still thinks like that; traffic growth is all you need to justify carrier spending. Profits are for the vendors.
When you look at initiatives like the cloud, SDN, and NFV, what you’re really seeing is a buyer community who feels abandoned by their sellers, pressured by market forces, and increasingly determined to make their own way in the world. The culmination of this feeling is the drive toward the use of open-source components for future network infrastructure. There has been operator interest in Open Source for five or six years now, but never with the management support behind it that we see now. Operators are rebelling against inaction in the most effective way possible, which is embracing the commoditization by embracing free components. Free stuff, after all, is always better for the buyer.
Ericsson got into professional services at the early part of the transformation shift, but they’re now facing the pressure of the new initiatives launched to disintermediate vendors. In particular, operators are suspicious of what seems to them to be a desire to build solutions for things like the cloud, SDN, and NFV one buyer at a time. They reason that if you want to support a network revolution you should have some product skin in the game. Not only does that prove your own commitment to the market, it lowers the amount of customization (professional services) and integration (more professional services) needed by creating packaged functionality instead of custom stuff. So Ericsson, lacking any conspicuous assets in the new revolutionary spaces, has a harder time pushing services.
Where this all matters to Nokia is in that point of “product skin in the game”. Nokia has 4G assets and not much else. Yes, we need 4G for mobile services but we also need shoes or tires, and nobody in the networking space sees those things as opportunities. Truth be told, mobile has always been subject to the same kind of commoditization as wireline, and that’s becoming clear in looking at the business behavior of the mobile providers. That means that doubling down on its mobile assets makes zero sense for Nokia, which means that the recent suggestion (by Nabila Popal, research manager at IDC MEA) that Nokia might buy Alcatel-Lucent is probably not a good idea. Similarly, despite the expansion announced in the Nokia/Juniper alliance for mobile cloud, buying Juniper for its products wouldn’t do Nokia much good. It just increases their exposure to a product area that’s commoditizing.
A more sensible alliance might be represented by the Alcatel-Lucent/NTT/Fujitsu announcement of a carrier-optimized (NFV-ready) server platform. The point is that if you want to do cloud, or NFV, or run SDN controllers, or host open-source software, you need to have servers. Cisco and HP, co-rivals to Nokia in the telecom world, both have servers and so they have automatic credibility in some of those key revolutionary areas. And that’s the important point. Nokia doesn’t need product representation, it needs strategic representation. With Alcatel-Lucent, it gets some nice assets but it gets an awful deep pool of bathwater surrounding the strategic baby. With Juniper it gets nothing, strategically. Servers are “strategic” for the future, but they’re already commoditizing (IBM sold off their x86 business to Lenovo, remember?).
Nokia, at this point, needs to become a strategic-something-company if they want to defend their now-retired NSN roots (which are about the only roots they have left). Their under-representation in hardware could be an asset if hardware isn’t where you need to represent yourself. But what software is strategic? Remember, operators want open-source stuff. Remember that Ericsson, who had a rather bad quarter, is the largest provider of OSS/BSS. The fact is that it’s going to take some careful positioning to find a niche where software value-add can be demonstrated, and sustained.
Is the cloud just a cheaper model for server consolidation? Is SDN just a way of pulling in white-box switches with lower margins? Is NFV going to cut the heart out of network-device differentiation by sucking out all the incremental functionality? If the answer to these questions is “Yes!” then Nokia has no good choices, nor do any of the competitors. Networking, as a separate industry, is doomed. Why? Because consolidation of network functionality onto servers will at least still demand servers. What’ in demand in the way of network equipment after such a consolidation? It would look like a giant ASIC. So add Intel or Broadcom to the list of possible winners.
You probably think I’m going to tout some new software area as the salvation of all these threatened players. I’m not, because it’s probably too late for any giant company to take a simple software route into profit. Nokia or any other vendor would need to become either a broad IT-like company—like IBM or HP—or they’d need to become a little software/services company. Buying somebody like Alcatel-Lucent or Juniper at this point only increases their risk. It would be best for them, and for Ericsson, and even for Alcatel-Lucent and Cisco and Juniper, to start thinking small, to consolidate around a core of truly valuable intellectual property, and grow out from it to cover product spaces that can be pulled through at a profit. Bigger traditional networking companies are only bigger dinosaurs, looking at that dazzling light in the sky.