We’ve all heard the phrase “Too little, too late”, and surely applied to telcos, but I want to propose another seemingly contradictory phrase to describe their current state. It’s “Too late, too little”.
Twenty years ago, telcos spent more per revenue dollar on opex than on capex (roughly 40 cents vs 22 cents). What I’ve called “process opex”, meaning the operations cost of the network itself (equipment and its support, customer support, energy, etc.) was 26 cents. To ensure that network upgrades made necessary by service evolution, standards evolution, and equipment updates didn’t make them look too bad on Wall Street, telcos worked hard to eat away at opex. Their particular focus was to cut staffing associated with support and consumer installations, and they made great strides in that area, resulting in cutting human costs by about 30% overall, achieved by 2020. Then they hit a wall.
Of 88 operators I’ve gotten comments from, 83 said that they had no overall plan for managing opex through that 20-year period, meaning that they’d kind of played whack-a-mole by tactical changes made predominantly to areas that were running regularly over budget. Of that group, 56 said that their reductions in human costs had been accomplished by measures that were only marginally beneficial, and all 88 said that their overall opex had not been radically altered. In fact, according to the operators themselves, they’d ended up with a process opex of 32 cents in 2024. Their measures had not kept pace with the factors that drove process opex, and opex overall, up. Why?
According to operators, the trends in consumer online usage and the business response to these trends is the biggest issue. Second is Wall Street’s demand for instant (meaning this financial quarter) gratification, followed by what some operators call the “tech news as entertainment” bias in user expectations. Then there’s unrealistic operator revenue expectations, followed by outmoded standards and technology consensus practices. This has pushed operators away from a thoughtful consideration of opex trends and of technologies that would best address them. Now, it’s too late, because there’s too little time and latitude to do the right thing. To see why, let’s look at each of our issues in more detail.
Online services, admit operators, are driven entirely by OTT players rather than by operators’ own plans. Every year, consumers consume more, businesses supply more, and everyone expects more because tech is worming its way into every life, every activity, more thoroughly. There is nothing telcos are worse at than opportunity or demand-side planning; they’re supply-side organizations through a hundred fifty years of DNA.
This has had a major impact on customer care. “The average household in our customer base uses more data than the average company did in 1980”, one operator told me, “and they’re just as strident in their protests against problems with their experience, but almost incapable of even basic self-support.” As a result, 29 of the 83 operators with no overall opex management plan said that their customer care in 2024 cost more than it did in 2000, despite sharp reductions in head count. In fact, 15 of the operators in the group said that cost rises were because of their headcount reductions. “Service-specific technology for support doesn’t survive radical changes in the service,” one said.
Why did so many whack opex moles, though? Wall Street wanted to see cost management that kept pace with necessary network capex budgets, and that precluded any opex strategy that took years to mature. Then there’s the impact of massive change. Why are air traffic controllers still pushing paper slips around to track flights? Same thing. Making a big change to a critical system can create critical disruptions, and from the perspective of the users (fliers in the case of ATC), the best you can hope for is that nobody notices what you’ve done.
Operators are reluctant to even estimate what a systematic approach would have cost, but the 11 who did offered an average of 60% increase in operations technology spending for a period of three years. None of the operators thought there was any chance they could have gotten approval from the C-suite for that, and none believed they could get it today either. “Yeah, we’re taking action. We’re sticking our head deeper in the sand,” one OSS planner said.
Of our 83 operators, it’s important to note that all of them said that vendors’ own interest in their Street creds meant that none were offering the tools for this vast transformation in opex thinking either.
The fact that both consumers and companies seek out sensational changes because stories on them are entertaining and get all the attention, and clicks win in today’s world, means that operators are also under pressure to respond to things that may be unprofitable or even impossible. Hedge funds drive stock prices because they drive trading volume, and they play on the PR side of new services and technologies. Hedge fund managers are not operations-literate, and in any event hedge funds profit from bubbles, and then again when those bubbles burst. You can’t offer boring-if-true comments on things like 5G, 6G, edge computing, or AI at an investor conference. You can’t offer exciting comments without spinning your cost wheels to deliver at least some trials.
All of this leads operators to accept optimistic and unrealistic estimates of new revenue opportunities. Look at 5G and what it promised, and look now at how little has been delivered. Of the 88 operators, 87 said that 5G was over-hyped, and 84 said that 6G was almost surely unrealistic as it’s evolving. Every operator said that many of their new-revenue plans of the last two decades were delusions, two-thirds said most were, and almost a third said that all had been.
Why are all these delusional things getting continued support? A big reason is the antiquated procedures operators undertake to define new services and their technologies, procedures honed when they were supply-side-dominated regulated monopolies. I’ve been involved in telco standards for decades, and I can tell you that any innovative and potentially revolutionary idea is inevitably buried. There are too many reasons for that to go into here, but I see no signs of this changing. Operators agree: “We look for consensus in a telco and vendor population that could never hope to reach it,” one told me.
What I find surprising is that only 9 of 88 operators hit on what I believe to be the biggest problem of all, which is competition and churn. For mobile operators, competition and churn are cited as the biggest factors increasing process opex; for fixed operators it’s still a major factor. When you offer a service whose only meaningful differentiation is price and reliability, and when the latter is generally considered “good enough” across all choices, you are forced to manage costs and try to offer suitable, effective, and cheap support.
The obvious question at this point, if you accept the views of operators themselves, is whether it’s really “too late”. Have operators passed the point where they could expect to remedy the situation? The answer to that, I’m afraid, is hard to pin down. They’d have to undertake a massive program of rethinking operations, a task they’ve been singularly unable to undertake in the past. In fact, they can’t even get internal agreement on whether to “modernize” or “replace” OSS/BSS, a debate I’ve personally witnessed.
That’s really the question here. In 2012 I was seeing operators face the question of a rethinking of operations, and dodging it. Nothing helpful, IMHO, has happened since. Was the topic so daunting it scared them off? I don’t know, but if operators don’t face it, they’ll have to face the consequences of falling profits.
