We have a bit of an interesting counterpoint story this morning, involving the future of networking and how we navigate transitions. Let’s open with Blackberry, who has announced they will be forming an activity to seek bids for JVs, acquisitions, or whatever.
Blackberry was the darling of the mobile phone space prior to the iPhone, and while it’s tempting to think that what won for Apple and lost for Blackberry was the coolness, the truth is that Apple won because they realized that the smartphone market would be won by somebody who could sell the phones to consumers. A mass market, when it exists, subsumes all the specialized markets, and iPhone success drive Blackberry out of its little niche.
I think everyone realizes now that the right move for Blackberry was to be more aggressive in the consumer space; in fact, they should have realized all along that this sort of thing would happen and even realize who’d be the player likely to make the move. Instead, Blackberry (RIM, then) hunkered down to defend its old position and withered there.
We now have a similar transition going on in networking. There was a time when a third of all network revenue was attributed to business services and when businesses were the only consumer of broadband. A typical T1 access connection (one end, not including the interexchange part) cost about nine hundred bucks per month. Now we have services in the hundreds of megabits per second for a fraction of that price. Networking overall has gone mass-market.
Business networking was about connectivity, and consumer networking is about experience delivery. Everyone knows that if you’re getting something delivered, all you care about is that it gets there in one piece, and it follows that experience networking doesn’t offer a lot of opportunity for differentiation for the delivery part. Bits are bits—ones and zeros, so differentiate that! But it also follows that if experiences are what you care about, producing them is where the money is, and will be.
This is what’s driving the business of networking. We are in an era where what matters is cheap transport and valuable experiences. That’s what the revolution to the real “next-generation network” will have to contend with. Produce bits as cheaply as possible and then toss them out like Johnny Appleseed to pave the way for profitable overlays. The telcos, like RIM, made a mistake by not realizing this and grabbing some of those profitable overlays when they had the chance.
For the network vendors, the people who make money pushing bits, we’re at a similar crossroads. The bits get cheaper and the experiences get better in our future, so you can try to capitalize on either or both of these moves. The embodiment of the trends that impact these two directions are the concepts we call SDN and NFV. Vendors have to accept that one or both these options will dominate their future, or they go the way of RIM.
SDN is IMHO a pure defense, a bet that commoditizing bits will provide a mechanism to retool networks to simplify operations and so reduce its cost, and also to specificize services to the exact connection model needed for delivery, tossing out the more generic architectures of IP and Ethernet communities that have dominated our thinking for decades. I don’t disagree that it could be something that helps some vendors gain market share, but it’s still a less-than-zero-sum game here. The bit-pushing pie is going to get smaller over time, more dominated by cost-based leaders like Huawei. If you’re honest with yourself you can see this already.
NFV is more complicated. Yes, the initial Call for Action white paper issued nearly a year ago was all about cutting costs by substituting general-purpose servers for more expensive appliances, but I think the founding operators knew all along that the architecture that hosted the virtual functions could rule the world later on, so to speak. NFV is the potential union of all the networking concepts of the modern age, all directed at framing that elusive NGN model. Whoever wins in NFV wins in NGN.
So who is that? A recent report suggests it will be Cisco and NSN, but I’m not convinced. Cisco does have the best inventory of NFV-related assets because it has (alone among the major network vendors) a strong server position. But even a strong server position isn’t a durable asset in an initiative that’s targeting commercial off-the-shelf servers. Cisco is going to get a RIM-like lead in the NFV space, an opportunity to shape its total vision based on a unique and early access to a subset of the market. We don’t know how to do NFV now, and somebody with all the pieces that could be used might well figure it out. But they still have to be able to exploit that knowledge in a durable, profitable way. And they have a big problem, which is that network transformation always hurts the incumbent the most. That’s why RIM thought they should stick their head in the sand with consumer smartphones.
Cisco shares that risk of overhanging old products with NSN. NSN’s “Liquid” stuff is smart positioning, but it’s not SDN and its not NFV, and if NSN were to make a big thing out of either of these technologies it would lead it down a path that would demand more network equipment, some of the very stuff NSN has been shedding in fact. NFV and SDN collide and cohabit in the metro, and NSN is more about mobile than metro.
Huawei is the Apple of networking, not because they have the consumer integration smarts but because they have the key asset in the network equipment space—cost leadership. All they have to do is not-lose and they win. For the rest of the vendors, the time has come to step up and take some proactive steps, before the Fate of RIM befalls them.