I know I’ve blogged a lot this week about industry conditions, but there’ve been a lot of developments in that space and we’re all in this together, industry-wise, in any case. Cisco’s analyst event offers more fodder for the discussion and also I think some clarity on the choices that face us.
At one level, Chambers repeated a theme he’s used so often that it could be set to music. He talks of market transitions and macro conditions and all the other good stuff, and I still think that dodges the real conditions. Look at the financial reports of Cisco’s customers. Industries are reporting record profits in many cases, and few are saying they’re starved for cash or credit. Yes, we’re in a period of economic uncertainty, but when in the last decade has that not been true to a degree? The thesis that Cisco (and, to be fair, most of Cisco’s competitors) seem to strike is “If things were just a little bit better, buyers who can’t justify spending more on networking because it’s not building a business value proposition would surrender and spend anyway.” I don’t think that’s realistic.
The Street has its own quirky views here, though. They want to claim that Cisco faces problems created by SDN. My surveys say that’s nonsense in a direct sense; there’s no indication that SDN adoption this year or next has any impact on equipment sales. What is true is that SDN has become a proxy for the real problem, which is that nobody will invest in something if it presents a dwindling return. If you ask, “Will SDN hurt Cisco?” the answer is “No!” If you ask “Will the buyer fears of dwindling network ROI both spur interest in SDN and in simply holding back on spending?” the answer is “Yes!” SDN isn’t the cause of Cisco’s malady; it’s just a symptom.
Chambers did address technology, even technology change. He talks about how Cisco is facilitating the cloud, partnering with others to help integrated solutions. But if you have a market that’s starved for ROI, you have to boost the “R” if you want the “I” to improve. Cisco like most other vendors in the network and IT spaces are offering guidance on how to do something with tech when buyers want to know how to do something different, something better.
The challenge Cisco has in the “betterness” dimension is twofold. First, their new market areas like servers may offer additional revenue opportunity, but they also offer (as UBS Research pointed out recently with respect to Cisco’s UCS) lower margins. The Street will beat Cisco up just as fast because margins are dropping as they will if sales are challenged. Second, even if Cisco knows what the buyer really wants and needs (which I think they do, and so do all their competitors, at least deep in their psyche) that doesn’t mean that it’s easy to offer it. Cisco is like all vendors forced to kiss Wall Street ass every quarter, and any sign of weakness tumbles the stock. That’s why I’ve been saying that we have a financial market that rewards tactical behavior. Well, anything to improve the fundamental buyer economics of network deployment is hardly likely to be “tactical”.
The most substantive thing Chambers said, the most positive thing, was that Cisco would be making a lot of acquisitions in the software space, not for “material” meaning P&L reasons. That implies they’re making them for strategic reasons. That’s critical because if we’re building the network of the future from SDN principles or NFV principles or cloud principles, how the heck did we overlook software? Cisco, to be a real IT company or even a real network company with a future, needs to be a software company and so do their competitors.
And the software isn’t a software replacement for switching and routing. OVS won’t change the game much, and “defining” the network is really about defining the network’s missions and services. Cisco is right in believing that it’s more important to get current network technology to slave itself to a software mission than to support legacy network missions with software instead of hardware. But they’re wrong in thinking that building ONEpk to allow for software control is enough. Without those higher-level service functions, without novelty and innovation in what networks can do for us, application control of the ONEpk sort are a bridge to nowhere.
That’s what we should be looking for in pronouncements by Alcatel-Lucent or Juniper too. I pointed out that Juniper, under its current CEO, was supposed to be turning itself toward software. It didn’t happen. Now that CEO and his senior software protégé are out and a new team is coming in. This could drive Juniper down the right road. Alcatel-Lucent is also promising a new focus, but their CEO isn’t promising that focus will be on software either. And unless you have a software vision in networking these days, you’re decoupled from all the trends you say you’re supporting. Live by the box, die by the box.