Gabriela Styf Sjöman: [00:00] The telecom industry still makes money out of a service we launched 100 years ago, voice. We’ve evolved it, but I think the challenge is that we focus more on the technical transformation, but I think that the most important right now is that commercial transformation, business transformation, because what we need to do now is evolve up the value chain. And that’s why I say we need to make more money. We save money, we put in those two buckets, but I think that commercial lens is something that we need to pay more attention to as telcos.
Announcer: [00:36] This is “Telco in 20,” a podcast that helps telco execs achieve a competitive advantage with AI and the public cloud. It is hosted by Danielle Rios, also known as DR. Today, we’re talking to Gabriela Styf Sjöman, the Managing Director of Research and Network Strategy at BT Group.
DR: [00:55] Hi guys, I’m DR. I was at a Telecom TV event earlier this year when I heard something that made me literally turn my head. A speaker was talking about AI and tech strategy, but instead of the usual jargon about speeds and specs, she said something eye-opening, “Every technology decision comes down to two questions. Does it help us make money or does it help us save money?” Business 101, right? The statement was brilliant in its simplicity. While everyone else is arguing about how networks have to have five nines of reliability, she cuts straight to what actually matters to shareholders, and she’s totally right.
[01:32] We need to ask what’s the business impact of our tech decisions? That speaker was Gabriela Styf Sjöman, Managing Director of Research and Network Strategy at BT Group. Gabriela is applying this thinking at BT, where she’s advocating what she calls commercial transformation. Today we’re going to dive into BT’s “AI for networks” and “networks for AI” approach, why best network marketing is dead and what comes next, and the cultural shift that needs to happen inside telcos to turn their teams into AI whiz kids. So let’s take 20. Gabriela Styf Sjöman is Managing Director of Research and Network Strategy at BT Group. Hi Gabriela. Welcome to “Telco in 20.”
Gabriela Styf Sjöman: [02:18] Hi, DR. Thank you so much. What a pleasure to be talking to you.
DR: [02:22] Well, we met at Telecom TV’s DSP event back in May, and you opened up a great panel that we were on, which was about AI in the networks. And the way you were speaking was the way that I wish all telco executives would speak, coming at it from the business perspective and not so much just technology. And that really piqued my interest and I connected with you, and so I’m super excited to have you on the podcast today. So let’s start with the fact that you’ve worked in 12 countries across three continents, and you’ve held executive positions at Nokia, Telia, Telecom Italia, and Ericsson before joining BT, the granddaddy of them all. And so you’ve been on both the vendor and the operator side. So which side is tougher?
Gabriela Styf Sjöman: [03:09] That is a great question. I would say they’re tough from different perspectives. I do think the network equipment provider, because they need to innovate probably at even faster pace. The margins are so slim. We have a number of structural issues in our industry that put a lot of pressure on them, and they just need to be on it all the time.
DR: [03:33] Yeah. I think most people have the idea that the grass is always greener, it’s easier on the other side, but that’s a super interesting response that you’re like, nope, I think it’s harder for the vendors, for the equipment manufacturers. So on that panel, you talked about a simple but powerful framework that I really like. Telcos need to approach technology strategy through two lenses, use it to help us either make money or help us to save money. I think this is business 101. And so why do you think so many telcos struggle with this basic business principle?
Gabriela Styf Sjöman: [04:01] Well, I think that there are multiple things. On the one hand, one needs to remember that the telecom industry does have structural issues. We provide the core of society. We connect people, societies, businesses, yet it’s not really necessarily valued in the same way. So yes, we’re challenged. Having said that, it’s still a profitable business. It’s less profitable in the past, but it’s still profitable. And I do believe that the mother of invention is necessity. When you really are in a crisis, although I wouldn’t want us to be in a crisis, I think sometimes that we haven’t really experienced that same pressure and therefore we still make money out of a service we launched 100 years ago, voice.
[04:41] We’ve evolved it, but I think the challenge is that we’ve focused more on the technical transformation, think about all these network transformation projects, digital transformation projects, but I think that the most important right now is that commercial transformation, business transformation, because what we need to do now is evolve up the value chain. And that’s why I say we need to make more money. We save money, we put in those two buckets, but I think that commercial lens is something that we need to pay more attention to as telcos.
DR: [05:12] Yeah, I think the airline industry went through a similar reckoning, where they used to brag about the planes. I have a faster plane or a bigger plane or a better plane, and then they had to really pivot to focus on the customer experience, to build loyalty. And that’s when we really saw all the loyalty programs come about. And now they price the seats different, like time of day, what URL you’re coming from, are you buying for a family, are you in economy versus business class? And so they’ve really innovated to continue to drive profits because it’s similar in that it’s a highly regulated industry with intense CapEx. And so how do you differentiate yourself? So you outlined BT’s tech strategy as “AI for networks” and “networks for AI.” Can you break down what this means and how they connect back to your revenue and profitability framework?
Gabriela Styf Sjöman: [05:59] Yes, and before I go into the AI, I think that this is the principle that we’re trying to apply to a number of technologies where we try to really distinguish between the strategy of what is a technology and how we are applying that technology. So we have, say, “AI for networks,” “networks for AI,” “Quantum for networks,” “networks for Quantum,” for example. So when it comes to AI, of course the first one, “AI for networks,” is to explore how AI can bring value for networks. And then we break it down and say the use of AI for intelligent operations, the use of AI for cybersecurity.
DR: [06:35] So is this AI more internal to BT, how you guys are changing your operations by using AI?
Gabriela Styf Sjöman: [06:41] Exactly. How do we apply AI to our networks, how we build it, how we operate it, how we deliver our network services. And the second one, “networks for AI,” means how do we build networks that deliver customer value and competitive advantage in a world where our customers are implementing AI for their operations?
DR: [07:05] Correct.
Gabriela Styf Sjöman: [07:05] That’s why we split them, because the one is the internal and the other one is, being a telco, how do we serve our customers and what do the networks need to be? And we’re working a lot in trying to understand different AI use cases, what language models, everything from large to small to multimodal, what capabilities will they need? What does the ecosystem look like? When do we think these will mature? And when do we need to be able to deliver certain capability in the network to serve our customers?
DR: [07:34] Yeah. And around the industry, you’re seeing lots of different applications of this idea. I think some people are chasing that GPU as a Service idea, partnering with NVIDIA or with hyperscalers to open sovereign data centers to support AI workloads.
Gabriela Styf Sjöman: [07:48] Yes.
DR: [07:48] I saw a few weeks ago, SK Telecom doing ChatGPT free for all subscribers for two months. I’m seeing some other telcos come up with some ideas on their enterprise side to help small and medium or maybe even mid-sized enterprises apply AI to their businesses. And so there’s a lot of different ideas. I don’t know what’s going to hit, I don’t know what’s going to work, but it’s a very big experimentation phase that we’re going through right now.
Gabriela Styf Sjöman: [08:15] Oh, there’s a lot. Like you say, nobody knows, but this is where co-innovation, partnerships and ecosystem play a very important role. We’re doing many things with our customers, with partners. We’re opening up all the time to exactly what you say, that exploration, because whenever I go to some conferences or I meet partners, there are those who often promote the explosion of data in the networks. We haven’t seen it. And then I try to correct people and say, “Well, there’s going to be an explosion of new needs within the data centers or between intra-data centers, but not necessarily across the full network.” So remains to be seen. It remains to be seen also what the telcos play in that value chain. Do we once again remain the pipes that just transfer all this data or can we play a more critical role?
DR: [09:06] Yep. Has it been hard to change the actual work habits of the people inside BT to start experimenting and using AI? How are you guys getting, how many people work at BT? I mean tens of thousands of people, to change the way that they work?
Gabriela Styf Sjöman: [09:23] It’s been a journey. We’ve been doing this, just like many other telcos have been doing this, for a long time. It’s not new and I think we have pockets of different levels of maturity. In some areas we’re doing a lot, and of course the first step is just to simply automate things that don’t add value.
DR: [09:41] Easy. Yeah.
Gabriela Styf Sjöman: [09:42] That’s easy. And it’s a no-brainer. I think where we now need to embrace a different culture, which might be more difficult, is how do you scale this beyond what I often call domain-based or component-based automation to more of a system-based. So we begin to talk about policy-driven slices, for example, or even intent-based networks that require more of an end-to-end programmable …
DR: [10:13] Experience.
Gabriela Styf Sjöman: [10:14] Yeah. You need to have that end-to-end view. That requires collaboration, but it also requires redesign of processes. Very often in this journey, we tend to underestimate that the teams need a skill set that we haven’t had to have for a long time, which is redesigning our processes. Many of the processes that we have in the telco world and in networks have been around for a long time. How do you reconfigure the network? And I think that’s probably one of the things is skill sets, not only in AI per se, but it’s the process redesign and of course the culture of collaboration. How do you bring that together? And above all, I think another thing is for the network’s organization to think, consumer service. We are used to thinking about domains, network resources, perhaps the network service, but bringing the skill set and a mindset that what matters to our customers is the consumer service. And that is where we need to understand how can AI help us with a strong customer experience that is much more tailor-made, on-demand and so on.
DR: [11:15] Yeah, I mean it’s just such a switch. I would almost describe it as switching from being network-centric to becoming subscriber-centric, and it needs to happen in the network group. And a lot of the time, those people are very much like, “Stay away. I’m doing hard stuff. Leave me alone. Send the pizza under the door. Subscribers, that’s marketing’s problem, not mine.” And changing the culture within the network groups, it’s going to be tough because they’ve been thinking this way for 100 years. The network at all costs and everything else comes secondary.
[11:47] And that leads me to my next question, which is telcos love to differentiate based on speed and coverage. For example, advertisements on the TV are things like, “We have the best network, we have the most coverage, we were the first 5G in the country,” blah, blah, blah, whatever they say. And so at some point though, and it’s happening faster and faster, we’re reaching network parity, which as much as you guys don’t like to hear that, can’t tell the difference in the United States, AT&T, Verizon and T-Mobile are, sorry guys, they’re kind of the same. And so how do you think providers are going to start to differentiate themselves in the future?
Gabriela Styf Sjöman: [12:19] It starts with the customer, always. In the past year, we’re maturing this particular discussion where we say, “It’s not about necessarily being the best network. It’s about the most competitive offering.” And the most competitive for me is about that value proposition, and that requires that you understand your customer. What do they want? What are they willing to pay for?
DR: [12:45] What do they value?
Gabriela Styf Sjöman: [12:47] What do they value? And in a world where the majority of the services that most people use, I’m not talking about the edge cases of industrial automation, the speeds are good enough. What’s more important is the reliability of the network. We want reliable service. I want to know that I have it all the time.
DR: [13:05] Yeah, I want it to work at my house, I want it to work at work, I want it to work on my routes.
Gabriela Styf Sjöman: [13:10] Exactly. That seamless experience, always on, and that seamless handover between different accesses, those things matter a lot. But we also go towards a world where not only do we now need to focus more on the service that we deliver and who’s consuming it and what they need, but also the customer segments. If you look at most telcos, we say consumer and enterprise, but there are a lot of types of consumers. So now I think the next big thing is to say we need to do better customer segmentation, better understanding where customers are and what they actually really need.
DR: [13:42] Well this area, I’m super passionate about. I totally believe hyperpersonalization is a font of revenue that has not been really tapped in for telco. The data that you guys have on your subscribers, the routes that they take, where they live, how many lines they have, that’s such an opportunity to optimize, like I said in the airline example, optimize pricing. And I get you’re regulated. I’m not saying go across that line, but there’s so much rich data that, because of the way that we’ve been doing our technology, is inaccessible, you can’t really access it very quickly.
[14:19] For my company, Totogi, I’m crazy passionate about doing this for telco. I’m like, you guys are constantly looking for all these new revenue use cases. I’m like, it’s right in front of your nose. You’re sitting on a pile of gold and we should just figure out how to start to monetize that. And so you’ve said that telcos need to stop talking about technology and start talking about products and services. Given that our industry loves specs and standards, what do you think is holding us back from making this shift?
Gabriela Styf Sjöman: [14:45] I think it’s a cultural change. Just go to any conference that we have and predominantly the panels are about the technology. And whenever there’s an O-RAN, for example, I’m selecting one area.
DR: [14:56] Yeah, you’re picking one, yeah.
Gabriela Styf Sjöman: [14:57] The panel will be about everything from standards to the vendors. Very few discussions are about the outcomes that we’re looking for and how O-RAN can help support that, as an example.
DR: [15:10] Correct.
Gabriela Styf Sjöman: [15:10] That’s why I like when you say applied. What is the purpose? What am I trying to solve for, and how can any technology solve for it? And I think this is, by the way, also going to be a big shift in how we view data. We haven’t really had a culture of purposefully handling data. How do I purposefully, with purpose, collect data, not just build a bulk collection. What do I need real-time? What don’t I need real-time? So I think this is also something that impacts across processes, how we approach customers, but all the way to how we handle data.
DR: [15:48] Yeah, I mean, I don’t think this is a capability problem. This is not a technical problem to solve. This is an attitude and culture change that we need to do, but we’re at a new crossroads with AI. It’s a new opportunity to embrace new technology that can help telcos, as you said, both inside the company, in making you more efficient, smarter, make better decisions, but also more business opportunities. I know for one, I am using more data. Yesterday on the plane I had Claude working on a blog, and I had ChatGPT working on some thoughts I have around Elon Musk and his spectrum buy, and I had them going at the same time. So I’m consuming voluminous texts coming back and forth as I refine a newsletter or work on a blog. And so behaviors are changing. It’s really interesting.
Gabriela Styf Sjöman: [16:35] You remember the last time we spoke, there was something I just thought was great. You measure how much AI your teams use. I thought that was great. I mean, what an amazing way to also incentivize people.
DR: [16:49] Totally.
Gabriela Styf Sjöman: [16:50] There is a little bit of a shame with some people, like, “Well, I’m supposed to be smart enough, I shouldn’t have to use AI.” And I said to people, “It’s an augmentation tool.”
DR: [16:59] It literally makes me smarter.
Gabriela Styf Sjöman: [17:01] Exactly.
DR: [17:01] My blogs are better. I have a sparring partner that I can bounce ideas off of, and yes, you’re right, we measure our employees on how much AI they use, and I get a report every Sunday night and what the percentage of use was across the organization. Who are my leaders, who’s making moves and who are my laggards? And I go to the laggards and I’m like, “Why are you not using AI?” We buy them all the tools, everything. They’re free to experiment.
[17:28] A new startup comes out, they can go test it out. I have this chat thread called the Water Cooler. We have a completely remote organization. And so you share tips, you share ideas. And so we use Cursor for programming, and Cursor came out with a really cute little bug finder that helps you find bugs in your code proactively. And I’m like, “Go use this, everyone, now.” But that’s how it happens. It happens at the grassroots level, it has to happen at the leadership level and then everywhere in between to change behaviors. It’s tough. So we have something in common. I am Mexican. You are a Swedish-Mexican. You have a Swedish father and a Mexican mother. And I found that fabulous and interesting and awesome. So my first question to you is is your favorite food an albondiga, which is a Mexican meatball.
Gabriela Styf Sjöman: [18:18] Oh, I know. I get nostalgic. My grandmother used to make albondigas.
DR: [18:21] Well I think the last time I had albondigas was with my grandmother. They’re hard to make, but they’re delicious. And so what’s your favorite Mexican food?
Gabriela Styf Sjöman: [18:31] Oh, there are many, but I like crunchy food, so I like flautas.
DR: [18:37] Oh my God, I love flautas.
Gabriela Styf Sjöman: [18:37] I love flautas and I love tostadas.
DR: [18:41] Yes. Well one day, Gabriela, I should make you my crispy tacos I make with a picadillo. Every time I make them, people say I should start a taco truck. And I’m like, “I’m kind of chasing my telco dreams. But if you think I’m going to do better in taco trucks, maybe I should go do that.” Well, maybe one day, Gabriela, I will cook some crispy tacos for you.
Gabriela Styf Sjöman: [19:04] I would love that.
DR: [19:06] But this is an amazing conversation about what’s going on at BT and what you think about AI, and I really appreciate the time. I know you’re a super busy lady, so thanks for coming onto the podcast.
Gabriela Styf Sjöman: [19:15] Thank you DR so much. It was a privilege talking to you.
DR: [19:18] Awesome. Stick around. We end each podcast with the “Telco in 20” takeaway. I’ve got two minutes to tell you something you need to know.
[19:34] Gabriela and I just talked about BT’s “AI for networks” and “networks for AI” strategy. And she nailed something most telcos miss. Getting value out of AI isn’t a technical problem, it’s a cultural one. At Totogi, we embrace AI fearlessly. We give our people access to everything. Claude, ChatGPT, Cursor, whatever, train them on data safety and set them free. We measure every person’s AI usage daily. For example, last month, our forward deployed engineers used AI to improve CR delivery productivity, three to 4x. That means faster projects, lower costs, and higher quality for you. So stop hiding behind the abacus and embrace the computer. Make AI part of your daily workflow. Measure it, reward it, and let your results speak for themselves.
[20:27] To prove we’re eating and drinking AI 24/7. I’m bringing one of our FDEs to Telecom TV’s AI-Native Forum in Düsseldorf next week to show everyone how we build real AI telco applications using Totogi’s BSS magic. No yapping, no slides, no demos, just real building. Come find me there or connect with me on LinkedIn and X @TelcoDR and we’ll set up a time to meet. Until then, tune into more “Telco in 20” episodes, like and follow, and leave us a five star review. Don’t forget to sign up for my no BS email newsletter on telcodr.com, and check out our awesome YouTube channel and hit that subscribe button. Later, nerds.
