Can Networking Control Its Own Destiny?
I remember a meeting at a bank in New York in the early 1990s. The topic was strategic technology planning, and the CIO and the EVP of operations were co-chairing. They were taking a poll on the views of the key technology leads, and the EVP turned to their network operations head and said, “We’ll start with you.” It was a clear statement of the role of networking at that bank.
It’s been 15 years since that meeting, and I had an email exchange with the new CIO of the bank, who was running down the list of people he planned to interview for detailed status and views. The new network operations person was not on the list. What changed?
What, indeed. The issue, it turns out, isn’t limited to our New York bank either. In the early 1990s, the strategic influence of network operations management on overall business planning was about the same as the influence exercised by senior IT management. Since then, the influence of IT has risen virtually every year, according to CIMI Corp. research, and that of network operations has fallen. Quantitatively, IT influence today is 21 percent higher than it was in that early 1990s meeting. Network operations influence in the same period has declined by a third.
Subjective input validates these numbers; I’ve heard this story from more than a few network executives in enterprises. What’s really fascinating is that if you ask a veteran network exec about the “good old days,” they all place the best of times around the same point, which is 1999. That’s the year my research shows was the peak of networking influence; when in fact, network executives were the major factor in business planning related to technology use. Few of those veterans remain on the job today, but those who do will verify that the “good old days” seem gone forever.
One of the most telling statistics out of my research is the relationship between the influence of network executives on IT planning, and that of IT executives on network planning. In 2003, an IT exec was about 1.6 times as likely to be consulted on network planning as the network exec was to be consulted on IT planning. Just four years later, it’s gone up to 1.7 times as likely. Compare those numbers to 1999, when the network exec was 1.7 times as likely to be consulted on IT planning as the other way around.
According to networking veterans, the problem is that the CIO is the chief technology player in an enterprise, and that person has, almost from the start, been an IT executive and not a networking type. Two-thirds of the networking executives I talked with said that CIOs were likely to discount their views because the CIO lacked network experience and “didn’t understand network problems.” Almost exactly the same percentage of CIOs think that networking executives don’t understand “the big picture.”
Others told me that it was all about money. IT budgets have increased substantially relative to networking budgets in the last 15 years, and the people who spend the most money have the most clout with management, or so some have said. The problem is that my survey data shows that the budget increase was a lagging indicator; it happened after the influence decline and was thus likely a response to lost influence and not a cause of it.
The word from the top seems the most aligned with the overall survey data. If you ask VPs of operations, COOs and CEOs why IT gets more influence than networking, nearly five out of six will say that it’s because IT has a more direct influence on productivity improvement. Anyone who’s fought the enterprise technology budget wars knows that ROI has been king for decades. In the post-bubble Sarbanes-Oxley era of conservatism and cost management, that’s more true than usual.
If we follow the trend line of my research forward, we’d find that by 2013 the IT executive would have three times the strategic influence of the network executive, and IT management would be consulted in network planning more than twice as often as the other way around. One disgruntled senior networking manager put it this way: “In five years my successor will be sharpening the CIO’s pencils, and we don’t even use pencils!”
The trend we’re talking about here is significant beyond the careers of the various executives. Where an advocate loses strategic interest, the thing being advocated loses budget power. Where senior management ignores details of a technology decision, the products being selected are harder to differentiate on anything but price. Where account control is based on a falling star, the vendor falls with it.
If you look at this picture, you can understand why Cisco is suddenly talking datacenter and software. A company that has historically relied on account control clearly has to be engaged with the executive who’s controlling the budget in the account. The recently enhanced cooperation between Cisco and Microsoft represents another indicator; Cisco is making sure that it has engagement with the company that is perhaps the key software player.
The trend also explains why some issues in networking, like security, seem very hot and get a lot of budget and a lot of attention. These are network- related issues that engage IT planner interest and senior management interest as well. So do things like collaboration and application-awareness in network monitoring and management. In fact, it seems as though practically everything about networking is interesting except the primary mission of networking, which is pushing bits around.
Service providers worldwide report that their revenue per bit is falling by 50 percent per year. It follows that bits are becoming a commodity, and bandwidth pricing for the enterprise shows the effects of the trend. Back in the early 1990s when that first banking meeting I referred to was held, the bank paid $290 per month for 64 kbps access connections to branch offices, not to mention what it paid for the long-haul connection. Today, the same bank gets business DSL for VPNs at $58 per month, and that gives them 3 Mbps bandwidth to the branch.
This means that the network operations plan to deploy a traffic concentrator that would cut bandwidth needs by 40 percent could have saved about a hundred bucks in 1992, and would save only 20 bucks today. The bank reports that the cost of the device to do this concentrating is about the same today as in 1992, so their ROI on the decision has fallen to one-fifth of its 1992 level.
Keys To The Kingdom
For the network executives hungry for attention, it’s pretty clear that application awareness, security, collaboration and stuff like that are the keys to the kingdom. Unfortunately, they’re not a lifetime pass.
Enterprise IT and networking management are proxies for their respective vendor communities in terms of technology direction, and the IT vendors are a lot more aggressive. Cisco has to defend its “software strategy” in the press despite the fact that this really represents only a nit of revenue, but every single major computer and software vendor has a networking strategy. EMC, for heaven’s sakes, has a VOIP monitoring offering, and they were supposed to be a storage vendor!
Unless we want the networking industry to start looking like the PVC pipe industry, we need to be asking Cisco why they’re not doing more with software—instead of asking them to justify what they ARE doing. We need to be asking other enterprise network equipment vendors, including Extreme, Foundry and Juniper why they aren’t jumping onto the same software bandwagon.
There’s a big demilitarized zone between the bottom of the application and the top of the network, and everything from collaboration to SOA fits somewhere in that zone. The stuff that goes to the network vendor goes to the network budget, reinforces the network’s role in strategy, and builds margins and revenue along the way.