Are enterprises worried about their network equipment vendors’ health? Do they have something to worry about? What’s causing whatever is happening, and how might it change networking? All good questions, and ones we’ll try to answer here.
There have been multiple stories about how things like the HPE/Juniper deal and the Cisco restructuring are making “enterprises nervous”, and in the second half of 2024, over 90% of enterprises I chatted with told me that they were “aware of” or “watching” these trends. Only 18% were “concerned” or “alarmed”, and even this group pointed out that any business changes one of their vendors made was always a potential issue. “I’d like all my vendor partners to have a stable model,” one CIO said, “but I also want them to evolve to serve market conditions and my own requirements.” Change, then, can be bad, but it can also be essential.
While the article I cited above notes an analyst view that some enterprises are putting “things on hold”, I didn’t get that comment from anyone, at least not referencing vendor changes. There are plenty of uncertainties in the market right now, and over the last couple months, and so there are reasons beyond vendor shifts to be slow-rolling some projects. I think we have to chalk up some of the “concern” to coincidental factors.
Not all of it, though. I think that if you look at the HPE/Juniper and Cisco moves, you can’t ignore the fact that they both portend shifts in the business model of network vendors enterprises depend on. The questions this raises, say enterprises themselves, boil down to two big ones. First, are changes being driven by things that could help, or hurt enterprises? What’s best for sellers isn’t always best for buyers. Second, are any of the shifts driven by something that could indicate a fundamental shift in technology or practices that enterprises should be planning for? Both questions are complicated.
M&A and restructuring in the vendor space is almost always associated with a significant change in market conditions, something that enterprise buyers aren’t always (or even often) considering proactively. My view, based on my own interactions with enterprises over many decades, is that the change in play here is the shift in network investment off the critical path of empowerment. Thirty years ago, there was plenty of information available to applications that could improve productivity and decision quality, but a shortage of means of connecting that to workers. That’s not true today. For the last three years, the number of enterprise complaints I’ve heard that network delivery of application/data access was limiting them has been zero. For the last twenty, it’s been consistently less than fifteen percent, and declining steadily.
If enterprises’ need for connectivity is being met, then they don’t need to spend more every year to accommodate change; there isn’t any. That means that network budgets are under pressure to do more for less, since every company (vendor or buyer) is looking to manage costs to improve profit. This is demonstrated by the steady decline in the growth of spending on new network technology over the last thirty years, and the explosive interest in new (however hypothetical) drivers of network change, like AI.
The impact of this on vendors is obvious. They’ll attempt to engage better on projects that actually do represent incremental opportunity, and that takes two forms. One is getting aligned with the influencers of those projects, and the other is shifting technology development toward the stuff the projects involve. HPE/Juniper is an example of the first, and Cisco’s reorg of the second.
Almost from the first, my interactions with enterprises have shown that data center changes drive network changes, and that vendors with the most strategic influence in the data center have the greatest chance of controlling opportunities in network equipment. Juniper, driven first and foremost as a provider of routers to network operators, thus falls short of having a “bully pulpit” in the data center. Years ago, in fact, I criticized them in their lack of aggressive positioning in data center and application evolution (see, “A Strategy of Absence or an Absence of Strategy” in Network World, perhaps still available HERE). HPE, next to IBM, has the most strategic influence in the data center, so they’re in a better position to leverage Juniper, providing them actually influence Juniper’s marketing/sales strategy. The HPE/Juniper deal, in my view at least, reinforces the switching area that enterprises spend most on.
With Cisco, the situation is more complex, at one level simplifying the organization by merging security and networking to recognize the potential interdependence, and at another introducing things like the cloud and AI as megatrends that cut across both. Organizing to focus on security doesn’t, in itself, commit to shifting development emphasis away from switching, but it doesn’t reinforce it either. The Cisco comments in the article I cited don’t do that either. ““Networking continues to be incredibly important to us and we’ll continue to support that space as well,” Herren added. “But it’s looking for efficiencies as we look across the company really in every way so that we can take those resources and allocate them into the fastest growing spaces.”
The common point here, between the HPE/Juniper deal and Cisco’s reorg, seems to be AI, but it’s really what enterprises said they wanted, which is support for change. Networking is stagnating in its traditional mission, but there are things that pose a risk to tradition, to the networking status quo, boring as it may be. AI and the cloud are two at the head of the list of those risks.
Enterprises don’t fear the changes as much as some might believe, perhaps thinking that they’d be asked to do more with less. Well over 90% of enterprises I chat with are sure that if major changes come about because of the cloud or AI, they’ll be projects that link new benefits to new spending, and those projects can kick in some network bucks too. Many welcome them, because it’s easy to see how new projects could help add interest to a CV, relieve boredom, etc.
Should they be more worried, though? Perhaps. It’s clear that the basic network equipment they depend on is commoditizing, and that could force some players to pull back from the lowest-margin pieces of the business. That in turn could mean enterprises have to validate new vendors, perhaps multiple vendors, and none of the ones I’ve chatted with are eager to do that. I think they’re safe for now, but out three to five years? Change is in the wind.
