Supply and demand shape our industry, like they shape pretty much everything in every market, and we see some signs of them both operating today. We also see signs that they often create long-term change after short-term disorder.
On the demand side, corporate raider Carl Icahn has taken a 10% share in Netflix, embattled OTT video player whose promise to change the game in video hasn’t quite worked out. There are a lot of people who will see this as an indication that the Netflix model is going to be rejuvenated, but Icahn’s investments are more likely to end up being dismembered than as emerging industry giants. The real question is where he thinks a buyer will come from, because a buyer is what he wants.
The obvious answer is the Big Three; Amazon, Apple, and Google. Handsets and wireless appliances started off as a kind of cellular-service camel’s nose, a way of attracting customers and reducing churn. Now the service is the tail of the appliance dog, and this is changing only in that handsets or tablets are at risk of becoming table stakes in a bigger war. I’m sure if you’ve read this blog regularly that you know of my views that all three of the Big Three are likely to become MVNOs over time. To build an MVNO value ecosystem they need something special, and it’s getting very hard to differentiate on appliances alone. What better way than to have a captive content play, something you can offer at a discount to your own customers?
On the supply side, we have news that Big Switch has raised another round in a victory for SDN—sort of. The real news here may be more that Big Switch absolutely HAS TO BE looking beyond its original OpenFlow model. No, I don’t mean just a “commercial” version of the OpenFlow controller. Frankly, that isn’t enough to create value in the real world (it might still be enough for the backers to flip the company; how much rationality is there in the world of VCs these days?) You have to think of SDN as a cupcake to understand.
Cupcakes are soft on top and hard on the bottom, and so is any realistic SDN model. If you want to be low down in the stack, at the OpenFlow level, you had better be a hardware player because anyone with a dozen programmers can field an overlay virtual network strategy these days—something Nicira-like. I doubt that Big Switch has been out there building fabrics on the sly, so I doubt that is where they’d go to find some value. For that role, keep an eye on Plexxi, a semi-stealth SDN player who has talked about their product enough to tell us it’s a switching concept.
So is Big Switch moving up, SDN-stack-wise? It has to, and the easiest nut for it to aim the cracker at would be the marriage of OpenFlow and cloud software, particularly Quantum and the DevOps stuff. You can’t do software-defined networking without some way of getting connected to the software, and at the least Big Switch has to defend this higher-level position against the risk that a player with a hardware solution will grab the icing and leave the dull part of the cupcake as the only game in town. No crispy edge, no sweet soft top, just…well…filler. But how many times have I heard startups talk about “laser focus”? In VC lingo that means “I’m going to give you starvation funding and so to accomplish anything with it you need to pick a micro-mission and run with it.” Big Switch has been laser focused on a little mission—OpenFlow control. Will the new funding give them the ability to expand, or just to primp themselves for sale?
These two news items encapsulate the complicated state of our networking market. We have profound business changes on tap because the guy with retail brand is the guy who owns the retail side of the food chain. There’s little chance that carriers can be that guy, which means that wireless will inevitably shift to an MVNO model. That focuses operators first on managing infrastructure cost relentlessly (bad for network vendors) and second on building service-layer features on top (also bad for them, given that the vendors have refused for five years to cooperate with that effort). The notion that the network is a slave to the features, which is what emerges from this, is the breeding ground for SDN. So, hopefully at least, our supply and demand trends may actually be converging.
That creates problems and opportunities at the vendor level. An MVNO craze would clearly shift everyone’s priorities, and quickly. That might favor the least formalistic of all the SDN players—Cisco. The Cisco SDN strategy has been a kind of double envelopment—support the cloud interface to own the icing and then provide a linkage to the crispy core that doesn’t really involve open protocols and standards (you blow those a kiss on your way past). The advantage of that is that it offers a holistic vision of SDN even if it doesn’t meet many of the formal definitions. The obvious solution, for Cisco competitors, would be to match Cisco in scope and embrace the standards processes.
The wild card in that is the Network Functions Virtualization stuff that I think marries well to the MVNO world…for carriers, handset/tablet players, and OTTs. A linkage between NFV and SDN (which the NFV people seem to be taking pains to say are only complementary) would make things harder for Cisco. Ericsson with its strong SDN assets and its OpenSAF initiative and Juniper (if it could retune its services vision for the Universal Edge and offer an SDN story instead of an SDN placeholder) could be in the best position to do that. We’ll likely see early in 2013 at the latest.
As a final point, I want to thank those who have read this blog in the years since it’s launched. Because of your continued interest and frankly your willingness to tell others, our October activity levels beat our record of 40 thousand hits per month! I’m gratified by your response, and I hope you’ll continue to let me know what you think!