Both Cisco and Juniper issued cautious guidance this quarter, which caused both stocks to take a hit. The issue, of course, is that “macro” conditions are said to be challenging, and that means essentially that buyers are slowing their orders, pushing back projects, not doing as much with network equipment as the vendors hoped, or even as much as enterprises thought they’d do at the beginning of the year. The obvious question is when the good times will roll again, and that’s a tough one to answer, even for the buyers themselves.
Let’s start with the network operators. Of 77 that I have data from, only 5 said that they believed their 2023 spending would be an uptick from their run rate through the first half. Not only that, only 19 said they thought 1H24 would be an uptick, and only 25 said they believed that their spending would tick up in 2024 overall. That’s hardly optimism.
The majority of the operators said the primary reason why spending wasn’t going to tick up was that the contribution 5G would make in spending has been declining since Q2. Most said the majority of their capital plans for 5G were “largely complete”, and there are no major budgeted projects on the horizon. Given that, they’re expecting to fall back to sustaining levels, which two-thirds say might even be a bit lower than the levels in the first half of this year, through all of 2024 and even into 2025. The 5G build-out budgets contained contributions for general improvements in infrastructure, and so that means that the useful-life clock of a lot of gear was restarted during the build-outs. It won’t expire for another three or four years, meaning the refresh of aging technology won’t contribute much to spending either.
What could change this? Thirty-four said that the only thing that could change their plans was a new service opportunity that offered an attractive ROI. Another 18 said that they might increase spending if there were a new infrastructure model, a new technology innovation, that would provably lower costs significantly. The rest couldn’t think of anything that would change their plans. However, even those who said that either or both these factors could change their spending plans didn’t see much change until late 2024. That means that the network operators themselves don’t see their capital spending ramping up in the next year.
How about enterprises. I have data from 164 of them, and their views are a bit more positive. In fact, 108 said that they believed their network spending for the year would be up 3% or more, and 146 said that it would be up at least somewhat. Enterprises were also more positive about 2024, with 119 predicting that their network spending that year would grow over 4%. Only 8 predicted lower network spending in 2024.
The downside here was that of the 108 who believed spending would be up, 99 thought that the largest component would be spending on network services. For capital spending, only 25 said they expected an increase in 2023 and 47 said they expected an increase in 2024. The biggest target of increased spending was security, and if you take that out of the picture you find that only 53 believed they would spend more in 2023 and 21 in 2024.
According to enterprises, the budget pressure is due to two factors. First, senior management has increased pressure to cut costs in order to boost profits at a time when sales growth is at risk. Second, new projects to empower workers or drive online sales or promotion are harder to identify. Cloud adoption, according to enterprises, has resulted in higher-than-expected costs and trying to optimize cloud applications is absorbing a lot of development and management attention. There’s also fear that there will still be a recession, and even fear that a budget impasse will impact US markets in the near term. In short, uncertainty reigns, and that never promotes aggressive spending plans.
Enterprises see the possibility that their plans could change, but their take on the nature of the changes is very different. The primary thing enterprises cite as a positive spending driver is a return to an easier monetary policy from central banks. They believe that the drive to cut inflation globally has resulted in a reduction in business spending overall, and caution on the part of the consumer. Enterprises who sell to consumers say that they’re caught in a vice—if inflation is high, consumers may spend less, and if central banks raise interest rates to curb economic growth, the consumer may also spend less. Thus, a return to “normal” network spending is likely to come when the global economy gets past both inflation and a recession risk.
Why, then, are enterprises a bit more optimistic on near-term spending? Unlike operators, whose 5G projects refreshed a decent amount of network equipment, enterprises had nothing to drive an early modernization of gear. They also have less gear to modernize, given that almost all enterprises use VPNs or SD-WAN and the Internet rather than building their own networks. At any rate, they are still modernizing, and one of the hot areas is what we might call “virtual networking”, meaning a combination of SD-WAN and SASE. These technologies are growing with cloud usage and in response to a need to reduce worker connection costs.
Of course, the ultimate goal of all technology projects for enterprises is empowerment. Work better, more efficiently, produce more, sell more, cost less. As I’ve noted in past blogs, we have had three waves of IT empowerment since the dawn of the computer age, and all were associated with a major advance in the way workers and information technology interacted. The network itself was, for a time, the limiting factor in empowerment. We had information, applications, and we lacked effective distribution of what we had. Now we have that distribution, and so the network isn’t the target of empowerment. What’s needed is another advance in the way workers and IT interact, and that’s complicated. It’s also largely out of the hands of network vendors and service providers. It’s an application problem.
Doesn’t that sound like the problem I cited for the telcos in advancing network services? We are outrunning the old software model, what I’ll call the “transactional” model, and we are likely heading toward a real-time event-based future. The question is what kind of software changes will be needed to support that evolution, and what applications will represent the low apples that can most easily promote it and justify it quickly? I think spending on the part of service providers and enterprises alike will depend on the answers to those questions.
