I’ve noted in past blogs that it was challenging to get network operators to embrace open-model technology. The primary concern they cite is “integration”, which translates to a fear that building a network from open components will expose them to risks of incompatibility and finger-pointing if problems occur. One interesting thing that came out of that earlier blog was comment from 48 of the operators, who wanted to tell me that they weren’t swearing off open-model networks, just setting parameters for their adoption. I think what they had to say is really interesting.
Let’s start with their big issue, integration. The problem here seems to be that the companies who offer open-model technology are first and foremost not typically firms they’ve dealt with in the past, and second are often rather small with limited resources. Unfamiliarity, in this case, seems to be what builds contempt, or at least uncertainty, and all 48 operators said that their “primary” telco equipment vendors wasted little time in subtly raising fear-of-the-unknown issues. They also, just as subtly, pointed out that small firms lack support resources and, if they fail to meet commitments, may be too small to present any meaningful legal recourse.
What operators want is either or both of two insurance policies. First, open-model technology from a well-known-to-them player, and second from somebody with deep pockets. According to 36 of the 48 operators, that’s one big reason why they’ve embraced public cloud providers for 5G support. However, 5G is a special situation in that there’s considerable need to deploy out of their home areas, where they have real estate, and because building out carrier cloud is a big scary investment.
What happens if they don’t get one or both of these? Then it comes down to a matter of trust magnification, and operators say that the only paths to that are to get an endorsement from another operator of comparable size, or to dip their toes into the open-model space with a smaller vendor to control their exposure while they’re building their confidence.
From these comments, I think we can see that there’s a fear factor at work more than a tangible dismissal of open-model stuff. In fact, of the 48 operators, 32 said that they’d had “discussions” with an open-model equipment vendor that could supply network devices, and 7 said they had engaged in at least a limited trial. However, five of those had supplied equipment for CTO-level lab trials that are notorious for not resulting in any significant deployment. Only two had actually done a limited field trial, but both said the results were encouraging so far.
Fears of this sort are surely the big problem with open-model adoption, but they’re not the only problem. All 48 operators said that one of the things that their legacy suppliers always brought up was the “time to innovation” issue. Our product pipeline, they tell operators, is full of neat stuff that’s going to roll out in the next year and that will offer you significant improvements in performance and operations efficiency. In some areas, like the often-cited 5G MIMO, it’s challenging to find open-model suppliers who are even current with technology levels. You won’t find any who have the R&D budgets and capital resources to innovate like us!
There’s an element of truth in this one, too, for two reasons. First, many open-model equipment vendors are in fact limited in R&D and engineering/production resources. Those that are still venture-funded are often told to spend on sales/marketing rather than on chasing new features. Second, open-model technology depends for time-to-innovate on the open groups that set the specifications. These always move slowly, taking literally years to advance something that a single major vendor could probably bring off in six months.
What’s interesting here is that only a few operators actually expected they might need these innovations in the next year, which means that their adoption wouldn’t be in the current budget cycles. The problem is that the legacy vendors pushed the innovation angle, which makes senior management nervous about it. The solution, according to operators is for open-model players to have positions on emerging technologies, a timeline for supporting important stuff, and a reason why they can’t yet commit to things that they don’t see as critical. “You have to show us you’re thinking ahead, following the trends,” one operator told me.
Which, to me, sounds a lot like saying “I need a reason to take an open-model risk and I’m not getting one.” Which raises the question of what the path to open-model success might be. Vendors who offer the open option seem to be confused too. Some think that you need to play up the capital savings, but others say that strategy has run out of gas, that those who could be swayed by it have already been swayed. Some think that you have to counterpunch the objections like integration issues or the problem of time-to-innovate, but others say that those objections aren’t real, that buyers who voice them are looking for an excuse to stay the course.
Then there’s a reprise on an old saw. Instead of “we have met the enemy and they are us”, we might say “we have met the enemy and they say they’ve joined us.” Companies like Nokia and Ericsson, with 5G technology, have agreed to some degree of open-model network compliance, a position I’m sure was largely driven by the desire to defuse what might become a major competitive issue. Some operators object to both these companies’ positioning themselves as “open”, but surely that move has given operators some cover. Still, both Ericsson and Nokia suffered sales declines this quarter and cited demand issues as the cause.
“Demand issues” means “prospects can’t make a business case.” This can happen for two reasons, whether you’re a telco or any other business. The first is that the actual return on investment is too low, either because of high costs or low benefits. The second is that the ROI target a project has to meet is set too high. This happens because there’s a higher perception of risk, one that the CFO feels must be countered with more financial conservatism. We have both these issues in play right now.
Open-model networks can raise the ROI. My conversations with operators suggest that the consensus is that capex for open-model networks is roughly 40% less than for proprietary, traditional, gear. That’s a major shift, enough to make almost any quasi-rational project look viable. However, this sort of gain is only meaningful when it doesn’t raise the perception of integration risk. According to 29 of my 48 operators, the risk of problems with integration and operations tends to be higher when only a small number of open devices are introduced into a legacy infrastructure. The remainder think that a few devices creates little risk, but as the numbers grow the risk rises until roughly the half-way point in displacement of legacy gear. Then it falls as the number of remaining legacy elements declines and so integration is less a problem. Either camp thinks that brown-field introduction of open-model technology is problematic.
We’re in an economic risk posture right now, globally at a high level and for telcos in particular because of falling profit per bit. Of those 48 operators, 41 say that they have more stringent business case requirements this year, and 35 expect even more stringent requirements in 2024. What operators who still believe in open-model networks hope is that some technology requirements shift, driven by opportunity, will provide the budget to write down more of the legacy gear and tip the balance to open-model networks. It could happen, but only if open-model players improve their time-to-innovate so they can actually exploit those new requirements.
