US telcos talk up opportunities in AI

In its Q3 2024 results this week, T-Mobile boasted of raising full-year guidance across the board after clocking its highest Q3 postpaid phone net customer additions in a decade, lowest Q3 postpaid phone churn in the company’s history, and hitting 6 million broadband customers. 

A section of the release went into how the firm is using AI to improve performance, but also plan network builds:

“T-Mobile’s combination of best network assets, customer centricity and technology leadership is expected to keep the company’s network years ahead of the competition well into the future. The company’s unique Customer-Driven-Coverage model employs AI based assessment of customer experiences utilizing T-Mobile’s network data in order to improve network performance, deliver higher customer satisfaction, and prioritize network investments where they matter most to customers.”

T-Mobile company execs elaborated on this in a subsequent earnings call, as reported by our sister-tile Light Reading. T-Mobile's networking chief Ulf Ewaldsson said: "It's an algorithmic methodology based on AI, where we're using billions and billions of data points that we are assessing from customer experience data across the network."

"We're correlating that with business data and with real customer outcomes – in other words, what customers decide to do on our network. And then we are assigning a CLV value, a customer lifetime value, to a grid across the country – more than four million little hexagons that we have created across the country, 165-meter-wide hexagons – we are assigning those values relative to competition to allow us to know exactly where we can build to please customers."

T-Mobile’s CEO Mike Sievert added: "We have tens of thousands of future projects that get ranked, based on some practical concerns like zoning and permitting, but mostly based on the outputs from our AI-driven algorithmic model called customer driven coverage. "And the result of that will be that we will stay ahead of the demand curve."

Verizon also posted its Q3 results this week, in which it reported 81,000 wireless retail postpaid phone net additions, compared with 51,000 net losses in third-quarter 2023 and a mild bump in operating income.

Verizon Chairman and CEO Hans Vestberg said: "This has been a pivotal quarter for Verizon, with transformative strategic moves and continued operational excellence. We continue to deliver strong results in mobility and broadband, and we are on track to meet our full-year 2024 financial guidance, with wireless service revenue and adjusted EBITDA trending at or above the midpoint of the guided range," said Verizon Chairman and CEO Hans Vestberg. 

There wasn’t much regarding AI in the press release, but on its subsequent earnings call, again reported by Light Reading, Vestberg went into where he sees some future opportunities for making money from AI: "Right now, we see large language models going to the big data centres.... As soon as they're going to be an application that you're going to use as an enterprise, you're going to put it much closer for the main reason of the transport cost for privacy, for security, and in some cases, also latency… Then you're going to see a big opportunity for us."

Head of the company's business division Kyle Malady added: "We're getting a lot of good orders from hyperscalers either on dark fibre or lit, and we're going to see that growing. But we have more than that, not just the fibre. It's the power, space and cooling, which you know is in really high demand. And we have a lot of latent assets in that area… So at the moment, we're putting it together. We're figuring out exactly how we're going to go into this market. It's a huge market. We can't cover it all, but there are certain segments we might be better off in than others."

We’ll have to wait and see what all this specifically manifests as – and every operator might not have the scale or capabilities to do what Verizon’s top brass appear to be mulling over – but it will certainly be interesting to see what the firm comes up with from a wider industry standpoint.   

AI could scarcely be more hyped as a sector, though it’s usually assumed the main beneficiaries of it in the coming years will be the Big Tech firms that have been busy ploughing resources into the LLM platforms. Much of the specific business opportunities remain a little ill-defined, but the prospect of making money from AI’s proliferation as well as consuming it for efficiencies will no doubt be a welcome one to telcos who are perennially on the look out for ways to boost revenues.