Tech Explained: – The Data Displaying AI’s Impact on the Telecom World in Simple Terms

Artificial intelligence is currently the shiny toy in tech, and when discussing it, most focus on the “AI stack,” data centers, or chips as the most vital aspect of furthering this innovative tool. Yet there is an invisible network that helps power this emerging technology and is just as vital to sustaining it; digital networks carry important data from place to place, and due to this, AI’s influence on the telecommunications world is undeniable. If AI becomes the defining application of the next decade, what’s the biggest strategic risk that carriers face right now? And who is building for the AI future versus simply reacting to it?

Roger Entner is an esteemed telecommunications analyst with decades of experience. Roger is the founder of Recon Analytics, where his work focuses largely on delivering near-real-time results for clients surveying the telecom, AI, and tech space. Previously, he served as a senior vice president for IAG Research, vice president for Ovum, and director of the Yankee Group’s Mobile Service Provider Practice. His well-rounded background makes for an interesting and engaging conversation.

Below is a lightly edited and abridged transcript of our discussion. You can listen to this and other episodes of Explain to Shane on AEI.org and subscribe via your preferred listening platform. If you enjoyed this episode, leave us a review, and tell your friends and colleagues to tune in.

Shane Tews: The competitive landscape in the telecommunications and cable space has been changing. Can you discuss what you’ve been seeing on your end through customer insight?

Roger Entner: The neat thing is that we observe the total market. We’re not focusing on one technology, we’re focusing on how people are using them. What we’re seeing, is interesting because the best solution for people, fiber, is not necessarily making them the happiest.

People who are using Fixed Wireless Access (FWA) and people who are using Starlink are happier in terms of Net Promoter Score than on fiber. There are several reasons for that. For a lot of people, FWA and satellite is good enough. They don’t have an alternative, right? Their current or previous provider is not treating them well and, when they have an alternative, they run for the exit. They’re much more satisfied that way. Also, both technologies are easy, especially FWA, and easy to use and set up. You buy the router, you take it home, it either works and you’re happy, or it doesn’t work and you bring it back. And we can’t find the unhappy people, and they go back being unhappy with their previous provider.

With satellite, you must buy the equipment, you must put it up, but if it doesn’t work, Starlink takes it back, no questions asked. That prevents these unhappy customers that are making or depressing their points. And so when we look at what’s happening, and you mentioned BEAD, and if we segment over there, satellite has been a big winner. Actually, Amazon has been an even bigger winner, undercutting Starlink sometimes by a third. They have been aggressive with this.

For very rural America, satellite is the right solution because you have low population density. You have relatively little interference from other people using it, so it’s great.

Then you have an area that when it gets more populated where FWA is great, right?

When we’re honest, does everybody need a gigabit of speed? No, they don’t. Is it handy when you want to download a game? But there’s almost nothing left that we’re downloading. We’re streaming everything, excluding games and applications. But everything else, we’re streaming.

4K video is 25 megabits per second. Multiply that by the number of members in your household who don’t want to talk to each other and watch a separate show. Then what do you need in the household? 50, 100 megabits per second? If that is rock solid, you’re good. And so that’s the secret of success for satellite and that’s the secret of success for FWA.

Then there are customers who want to have something more reliable, who want to have higher speeds. And for them, fiber is great, cable can be very good. It’s marrying the right technology with the right usage in the right location with the right customers is the answer.

Is the emergence of an AI divide as well as a connectivity divide starting to make a difference in the decisions the consumers are making on how they get connected?

We have the biggest database out there on AI usage, totaling 130,000 people right now and we’re adding like 6,000 every week. We asked all the AI respondents, who’s your home internet provider, who’s your mobile provider, where do you live? By asking them various questions we can tell what choices they made and then we can combine this then with a broadband map, the FCC’s broadband map.

When you set up your databases correctly and you have massive sample size, one of the neat things that can happen is that you can move from correlation to causal inference because your datasets become so large that the control groups are naturally occurring.

And with that, we can see several things. That there is a mutually reinforcing loop and causal loop where people on, say, fiber are using much heavier workloads with AI than people who only have access to DSL or FWA.

And at the same time, they are seeking out higher technologies. You have this mutually reinforcing circle where fiber is really driving AI usage and AI usage is driving fiber because more people are getting fiber where they have the access to that, which is fascinating. Both speed and latency are important because the way we interact with AI is very conversational.

If fiber is such a powerful accelerator for AI adoption, how much lag is there from the change to what we see when the behavior of consumers can catch up?

What the operators are rolling out is something called L4S, like low latency, low loss. Comcast, for example, rolled that out on cable. T-Mobile rolled it out on FWA and on mobile because it makes everything faster, more agile, and much more reactive. And when you don’t have to wait, it becomes much better. On an underlying basis, you know when this thing goes faster and reacts faster.

And so, you know, when you don’t frustrate your customer, they are less likely to leave. For cable, it’s predominantly a defensive move where you don’t want to aggravate your customer.

The happier my customer is, the less likely they are to leave and the more likely they are to bundle. I always say, happy people bundle. Bundling doesn’t make people happier. Because if I’m happy with what I get, I’m willing to spend more with you. If I’m not happy with what I get, I’m not going to ask for more, right? More unhappiness.

One of the key aspects and key drivers, especially for companies like Comcast or Charter or all the other cable companies that are under a lot of pressure is how do I make my customers happier? And then how can I upsell them to wireless where they make a good margin and the customer becomes stickier? And so that’s that voyage that helps. In the end, everybody is served because the customer gets a better product and enjoys it more and the company can keep their customers.

Can you give us the lay of the land for satellite moving forward?

Satellite is one of those very clear use cases where you have a very clear customer benefit. For example, I now have coverage where I didn’t before. And that is very highly beneficial. Starlink offers it today. Early next year, they will roll it out. They will bring full speed data, video, you name it. But again, this works really well where nobody lives. But you’ll be able to climb up a high mountain and take a video, share it with your friends, do a live stream on Facebook, and then you can call Mountain Rescue to call for the helicopter that rescues you and brings you down, right? All brought to you with Starlink.

It might be expensive, but it’s a phenomenal technology. What’s really, really interesting are the rumors that show that Starlink wants to become its own communications provider. They filed for a Starlink mobile trademark.

They have an MCC, which is like the international calling code that they need and it’s one that works in every jurisdiction. I wouldn’t rule it out. And he [Elon Musk] mentioned in some podcasts that, know, push comes to shove, he will buy the spectrum. We’ll see if he buys more Spectrum. There is Spectrum available. There are lot of opportunities for him to do this. Will it ultimately be successful is a completely different question because…take Dish with its network, for example. It is a phenomenal network, a technological marvel, and they were completely inept in selling this thing, right? Just because you have a great network, and just because Starlink might have a great network, doesn’t mean it will be able to compete selling it to consumers, especially when the customers in other places with other carriers are already satisfied.