The heart of a cloud is the data center, obviously. The heart of a cloud network is therefore a data center network, and as network equipment vendors struggle for profit growth, data center networking looks very promising. Given that, it’s not very surprising that Nokia is taking a position in the space. The question is whether it can turn a position into a win, and it won’t be easy.
Network vendors have been under profit pressure, in no small part because network operators and enterprises have been under pressure themselves. The one bright area of growth has been the cloud space, and as I noted above, the data center network is the centerpiece of any public cloud network-building plan. The same, by the way, is true for enterprises; my own research for well over two decades has shown that enterprises tend to build their networks out from the data center. That gives those vendors with a data center position a big advantage.
The “well over two decades” point is critical here, at least for Nokia. The position of the data center network has been known to the smarter vendors for at least that long, and players like Cisco and Juniper have been focusing increasingly on switching revenue for a very long time. Nokia is surely coming from behind, and that’s the biggest reason why their move into the space is a risk.
Every market tends to commoditize over time, with the pace of commoditization depending on the extent to which new features and capabilities can be introduced to justify additional spending. Data center switching is surely less feature-rich than routing, and routing is already commoditizing, so it follows that the switching space has been under price pressure. In fact, Facebook has been pushing its FBOSS open switching for five years now, and “white-box” switching was a big feature in the ONF’s OpenFlow SDN stuff.
There’s also the problem of sales/marketing. First and foremost, there has never been a carrier network equipment vendor who was good at marketing, given that their buyers were awful at it. Nokia is no exception. Then there’s the fact that to market effectively, even if you know the general principles of marketing, you have to know a lot about your buyer. That intimate knowledge is going to come only from direct relationships, meaning sales contacts. If you don’t call on a given prospect, you’re unlikely to know much about them, and if you’ve had no products to sell them, you’re unlikely to have called on them. Is Nokia a household word among cloud providers? Hardly.
This, for Nokia, sure looks like the classic long pass on fourth down in the last minute of a football game you’re losing. What in the world would make them go for it, other than that very football-game-like desperation? There are two possibilities, and making either of them work will demand Nokia make some big changes.
The obvious possibility is that Nokia is indeed in that last-pass situation. They’re behind both Ericsson and Huawei in the mobile space, and 5G is the only hope for carrier network equipment. The cloud providers are an opportunity, but so are large enterprises with their own data centers. Rivals Cisco and Juniper have enterprise sales of data center switches, and that gives them an advantage over a rival who might focus only on the cloud providers and operators. Could Nokia be looking to get into the switching space more broadly? Maybe.
The other possibility is that Nokia is reacting to the significant developments in the carrier cloud space. Network operators are committed to virtualization, and in today’s world the commitment is visible both in NFV in general, and in virtualization in 5G in particular. Future opportunities like IoT seem almost certain to demand hosting, and so it’s long been said that carrier cloud could be the largest single source of new data center deployments—including by me. The problem is that the carriers themselves have been extraordinarily (even for them) slow in developing any real carrier cloud plan, much less a commitment. They’re now increasingly favoring outsourcing of at least the early carrier cloud applications to the public cloud providers. Could Nokia see that outsourcing as a foot in the public cloud door today, and also see a future push by operators to return to hosting their own carrier cloud apps? Maybe.
If Nokia wants to be a broad-market data center switching player, their biggest challenge is that they don’t call on enterprises today and have little or no name recognition. To succeed, they’d need an incredibly well-done marketing program, with a great (possibly dedicated) website presence and great collateral. Without this, the burden placed on their sales personnel would be impossibly large, and it would be difficult to compensate them for the time needed to develop accounts.
Targeting only the public cloud providers might make things just a bit easier from a sales-marketing perspective, but this group has been looking more and more toward white-box switching and price pressure on the deals would be intense. Because the cloud provider buyers are quite platform-savvy, Nokia would need a highly qualified sales support staff to handle questions and step in when needed.
The last of the options seems best, at least in terms of opportunity versus effort. Nokia is knowledgeable about the carrier cloud opportunity, more so than nearly all the public cloud providers, which means they actually have an asset they could push through sales/marketing channels. Nokia, as a credible 5G player, has the same sort of positioning advantage in direct sales of carrier cloud infrastructure, so they could credibly tell a story of transition—start small with a public cloud using Nokia-approved technology, then migrate to your own clouds—to operators.
If all Nokia had was switches, this would be a lost battle for them. They do have more, however, including what they call “a new and modern Network Operating System”, a version of Linux designed to take advantage of microservices and cloud-native operation to unite a system of data center switches. This, obviously, could be extended to other devices at some point, though Nokia isn’t saying it will be. This makes Nokia’s story a story of cloud switching, which could be compelling.
“Could be”, but in order for the story to deliver, it has to overcome that sales/marketing challenge. Nokia, as with Alcatel and Lucent, has always had a very geeky approach to the world, one that hasn’t so far been able to overcome the basic truth that no matter how smart you are, you can’t win with a strategy that assumes your buyers are smart too.
Smart switches may be necessary conditions for a smart cloud, a cloud that’s capable of being elastic and efficient and operationalizable, but they’re not a sufficient condition. The cloud is an ecosystem, and you can’t optimize it by tweaking just the connective tissue.
The good news? Sales/marketing problems are relatively easy to fix. All it takes is determination and a good, systematic, approach. The bad news? Through the evolution of three companies, Nokia hasn’t been able to fix them. We’ll see how they do this time.