Automating the Digital Universe

22 Feb 2021
HeaDSpin team

HeadSpin proudly presents Converge: a show about experience, resilience, inclusion, and acceleration in the digital space.

In our very first episode, you’ll be listening to a keynote address by Srini Gopalan, Board Member for Germany and Managing Director of Deutsch Telekom, on how the telecom industry will evolve over the next few years and provide the foundation for the flywheel effect expected as other industries accelerate their Digital Transformation. He’ll first be introduced by Rajeev Butani, CEO of HeadSpin. This session was recorded live from our virtual event Converge on April 20th 2021.


Rajeev Butani: It’s my pleasure to invite Srini Gopalan on the stage. Srini is the Managing Director and Chief Executive for Deutsche Telekom in Germany. He is a dear friend and a colleague. I’ve learned a ton from Srini over the last two decades, I’m dating myself. Srini has been a true visionary and continues to be a true visionary across industry segments. He’s had significant impact in the consumer products industry segment, just to name a few, in financial services, and now he’s radically transforming the telecom area and looking at the digital transformation.

He’s leading Deutsche Telekom’s Germany business, and then through that he is driving grassroots transformation for the company, for the industry, and then for all of the businesses that are going to benefit from what Deutsche Telekom will do over the last many years. So please join me in welcoming Srini onto this. If you guys give us a couple of moments, we will transition from Srini and then over to you Srini. Welcome to HeadSpin and welcome to Converge.

Srini Gopalan: Thank you, Rajeev. Thanks, everyone for coming into this. My name is Srini. I run the German business for Deutsche Telekom, and I wanted to spend some time today talking about the future of telcos in a digital world, and what that’s doing to our business, our industry, and obviously some implications for all our partners. But maybe a good place to start, is to actually talk about the multiple layers within a telco. When you start with the physical infrastructure layer, on top of which we typically layer our connectivity technologies, and then the software architecture that sits on top, and finally the products and services that ultimately get delivered to our customers and clients.

For many years, this has sort of existed. These four layers have existed as a single monolith across either stationary fixed technology or mobile technology, where the telco has combined all four elements, so the physical pipes, the IP technology sitting on top of it, the software sitting on top of that, and then the services and the products in terms of voice or packets of data that come out of the other end. They’ve typically existed as one single monolith. To be honest, we’ve not spent a huge amount of time as an industry historically, decoupling these various elements.

But the one thing that I’m convinced off is over the next few years, we will see this monolith get significantly fragmented. At the physical infrastructure layer, we’ve moved from a single telco, whether fixed or mobile supplying the physical infrastructure, now to multiple different technologies coming in, whether that’s at the extreme satellites or even regional fiber companies (and you see a lot of this on both sides of the Atlantic) to alternative technologies coming in to the B2B environment — campus networks, very specialized networks for factories, to the infrastructure specialists, whether that is cloud companies, or the IoT specialists. As these new types of physical infrastructure have come in, the connectivity technology they’ve put on top of that has changed as well. But probably most importantly, the software architecture — which is what I’m going to spend a bit of time talking about, which is the key layer in many ways, because it connects the physical pipes to the end consumer product — that software architecture has become critical. Because what that software architecture has gone from doing is from running a single set of pipes, which is how we’ve typically seen the software of a network. It’s become much more an orchestrator: an orchestrator across multiple different technologies, and multiple different infrastructures. This is only going to accelerate.

I think the interesting thing from a telco perspective is you will see a fair amount of value migrate to this orchestration layer, migrate to this layer that really brings the different physical infrastructures together. Let me give you an example for a B2B customer, for example, who wants to be connected across the world. What they’re interested in is for someone to orchestrate the multiple different technologies that will be used across the world into a simple dashboard that allows the B2B customer to choose in which pipe they want, what kind of latency, what kind of speed, what kind of security. So that network orchestration layer is where more and more of the value of the industry will start shifting.

Then you have the products at the top end, which is what we provide our customers. Again here, you’re going obviously from a very simplistic world of voice and data to a world now where through multiple forms of communication — whether that is Zoom, Facebook — voice and data have kind of come together, especially as we’ve gone through things like the Corona pandemic, with more and more working from home, you see a coming together of traditional telco communication products, with things like WebEx, with things like Zoom, etc, etc.

When you step back from a lot of this, what you’re talking about as a telco is one fragmentation, but probably more importantly, a value migration from pure physical infrastructure to value sitting at the software layer, with things like connectivity technology and software architecture, actually just coming together as a single software layer, which both orchestrates and controls your network. That value migration is probably the biggest thing that we as telcos need to learn how to work with.

The way I think of it is: as a telco becomes much more software-led, effectively you get to the phenomenon of a telco as a platform, a platform which northbound opens up to allow different interfaces (classic examples being Zoom, WebEx, etc), but we’ll talk more about things like software defined networks. Southbound, it integrates with multiple different partners, assets, capabilities, and what you have is a middle layer, which is the telco as a platform, which brings together things like network orchestration, but also provides both northbound and southbound data and analytics, and probably really importantly, provides security, both north and southbound. That’s kind of what I mean broadly by platform services as well.

This change of what a telco is, from the fundamental source of value being ownership of the physical infrastructure, often protected by regulation through things like spectrum use rights and mobile, or simply the fact that there’s normally only one fixed network in any area, up to a value migration to becoming a platform with both northbound and southbound integrations I think is either a huge risk if you want to stay a traditional telco or a big opportunity if you’re a telco that’s willing to embrace the softwarization of the industry. Embrace the move we’re seeing to really a re-architecting of our production model to be much more cloud native or to be much more microservices and API driven. Most importantly, to be much more open to being a platform, and I know for a lot of you folks that’s a pretty natural thing, and you see being the platform as being a business model in itself, but for us telcos that’s something that is going to take some getting used to and certainly a fair amount of transformation in terms of how we build and market our services.

To give you some more sense of the service layer, we’ve talked quite a bit about the software layer. On the products and services side, what we see — and I’ll start with the right hand side on connectivity — as I said earlier, you’re moving away from a dedicated private enterprise network to really sliced networks. Let me give you some examples of that. As a car moves from being a piece of hardware to really a disaggregation between the hardware and the software — where a Tesla or a Daimler wants to do millions of over the air updates to its cars — you can’t any longer run a dedicated private enterprise network. There’s a lot of dedicated enterprise private networks for basically one long pipe connecting one factory to one central head office,

What you’re running instead is a really distributed mobile network, where some of the updates you’re happy to go over the internet. Some of the other critical security updates, for example, you will want a dedicated VPN or a dedicated secure transfer or a zero trust network, and you will want that on mobile. That kind of comes back to this concept of security as one of the product offerings becomes critical. Going back to my earlier sense of the telco as a platform — that platform northbound into B2B products has to be able to offer connectivity in the slightest form, has to be able to offer security (and we may or may not as an industry do the security ourselves, you may have security plugins that you open up to), and has to offer communication, and that communication will change from traditional telco voice and data to obviously a whole bunch of multimedia solutions. Hopefully that gives you a sense of what the northbound product and services layer lands up looking like, which are very different platforms with a lot of third party assets and a lot more cloud native in terms of the nature of its integrations.

Similarly, on the consumer side, again what you will see is a lot more context-aware, a lot more use of real time data. So again, let’s talk this in terms of examples. So with the gaming industry, there’s huge value in knowing — especially if you’re having multiple gamers online at the same point in time playing each other — there’s big value in knowing what the speed and the latency of the network is at any point in time. Being able to customize your buffering, your loading, the nature of your game to enable that gaming amongst two players, each with different latency at different points in the network, or being able to secure connectivity that assures you latency of less than x across all points on the network.

Clearly, that’s a much more complex network than the one we use today, where you know, you get broadly the same speeds or location on any device that you’re in, and the same latency on any, on any device at any low at any one location. Again, that’s a network that lands up being much more controlled by software, which dictates and orchestrates what different slice of the network you happen to be on. Another good example is we will effectively be running in the future two or three different networks at the least. Let’s say you’ve got a SIM inside a lift which monitors an elevator which monitors predictive maintenance. You could afford fairly low speed communication, and you don’t really care that much about the latency.

On the other hand, if you had a pacemaker and someone, and you wanted communication with data transfer to a Central Hospital location, you’d want very different bandwidths there and very different latency. Again, this goes back to the heart of the network that lands up becoming the network orchestrator and your ability to create the softwarization that so that you can then disaggregate the various slices of the network from the actual controller or orchestrator of the network.

When I pull all of that together and start giving you a sense of how we at Deutsche Telekom believe this industry will evolve — we believe starting from the bottom physical infrastructure, it will still be critical for us to own physical infrastructure, and there’ll still be value, especially in owning the last mile. The two areas of connectivity technologies and software architecture will be critical, because these will be the layers that will orchestrate the network, provide security in terms of things like zero trust, provide different slices and the ability to have different networks for two individuals or two different use cases. This is also the layer that will integrate both northbound into the product and services and southbound into the physical infrastructure. We’ve talked about products and services, and I’ve given you some examples of what that looks like in the context of both the consumer business and the B2B business.

The one critical thing that ties a lot of this together that we are obsessed about is that this is also a fundamental change in the nature of the organization we are. Until recently, most Telcos have been built fundamentally as builders of physical infrastructure, which is an industry which involves long paybacks, which involves a lot of planning, which involves long lead times, and a lot more you’ll have a certainty or tools built to manage long term uncertainty.

As all of you know way better than we do, the software industry is fundamentally different. The small investments are a different culture to an industry used to looking at 10-15 year horizons to do its planning. So as we talk about the the winners in this industry being people who learn how to softwarize the stack, so that for partners like several of you in this room, we’re able to provide the right micro services and API’s, both on the product and services side, but also on the physical infrastructure side. The heart of the Telcos who manage to do this will also be a cultural transformation moving away from infrastructure providers to being genuinely a combination of both infrastructure as well as software.

Like most things cultural are, it is a lot easier said than done. That said, we believe strongly that this is the destiny of this industry, and we believe strongly that we can be leaders in making this happen. That’s really the focus of where we’re building the company, but hopefully that also gives you a sense of a helicopter view of where we believe this industry is going. So with that, I’m going to pause and open up to questions.

Rajeev: Srini, thank you so much for this massively insightful thoughts and point-of-view? So we got a few questions coming up. I’ll probably start with the first question that’s been very dear to my heart over the last many, many years of infrastructure transformation that telcos have gone through: How will the telcos in this case stay prime and capture value of all of the transformation that’s being done versus this value being captured by the rest of the players in the ecosystem?

Srini: That’s a really good question, Rajeev. I think I’m reasonably clear on what we shouldn’t do. I often say we spend more time in software envy than in actually developing software. Right. On a more serious note — a couple of things that I think are critical. One: the physical infrastructure still remains important, and we can’t walk away from that, so we need to mention make sure that we continue. On the software side, I think the biggest thing we can do to ensure that we get our fair share of value is both open up and get better at it ourselves. Because I think the thing that doesn’t work is to try and believe that someone else will steal our lunch, and therefore let’s stay a walled garden. The reality, at least in the software environment, is the more open we are, the more we open up our API’s, both northbound and southbound, the more we’re willing to collaborate and work with people, the more we learn, and the better we’ll become at it. If we feel threatened — and this is what I meant by saying software envy — if we feel threatened and try and keep people out the ecosystem will only be smaller, and we’re not going to be able to do the cultural transformation I talked about.

Rajeev: That’s awesome Srini. This probably goes back to the next question that Manish Subramanian is asking. What are the new partnership models? You touched on the partnership models as a part of your presentation, but when you’re looking at this new landscape and new operating model, what are the partnership models that you see coming around, and what would be important for collective success of what we’re going through?

Srini: I think first one big part of collective success for me is how do we enlarge the size of the pie, right, which is not look at how we grab share from each other, but how we genuinely expand the size of the market. Because I think as Telcos softwarize, the opportunities for our industry as a whole are substantially larger. Then the question becomes how we work with people to actually create more opportunities rather than grab share from each other. Classic case of this for me is things like as cars become more softwarized, there is going to be opportunities for us as people carrying data, but there’s also going to be opportunities for providers of entertainment. There’s going to be opportunities across the board as a result of that.

I think on the partnership models, again, I think we as an industry have to be a lot more flexible around the types of things we’re willing to do. To give you a sense, we now have everything from a venture capital arm, which invests in startups or which invests in people whose technology we think is interesting, to in some cases revenue share, and somewhere in between, which is us being a distributor. I think we’ve got to be flexible about which model lands up working for whom. I know both of those sound like a bit of motherhood and apple pie, but I think this expanding the size of the opportunity rather than fighting for share and not being religious about what models will have versus not are critical to it. When I look back at where we as an industry dealt with smartphones, we didn’t do those two, which is why I don’t think we totally have participated in the smartphone boom to the extent we could have.

Rajeev: Right. You know, Srini, one of the things that we discuss internally is exactly to this point. I think it’s amazing for each of the players to play on their strength and take the right position on the field. One of the things that we’ve been debating internally with a few telcos is: how do the telecom companies and carriers take prime position in understanding the value and demonstrating the end customer value that’s being created across the ecosystem, because you’ve got the infrastructure, and there’s a whole host of innovation that’s going to take place in the middle.

There are a number of companies — if you look at history, and if you look at history as being a prediction of some level of prediction of the future — there will be Googles and Facebooks and a whole host of companies that will come and create amazing value propositions for consumers, and then you’ll have similar similar instances of companies coming and creating amazing value propositions for business to business companies. I think, in this case, how can telecom companies lead prime in understanding the experience and  being able to share what’s the experience that’s being delivered across the value chain, given that they control the infrastructure might be an interesting position to play in the field across.

Srini: I agree, and Rajeev I think a big part of that has to be to some extent our willingness to cannibalize some of our own legacy and incumbent revenue, and believe that actually moves us forward. Software defined networks is a classic case in point where in some cases, Software Defined networks will land up meaning us cannibalizing our existing MPLS networks — traditional Point to Point connectivity networks.

I’m convinced that’s a very short term view of the world, because if with time, the customer believes that some data can be carried over the internet versus needing a dedicated pipe they will find a way of doing it. If you can find a partner who can enable that quickly, then you’re better off doing that than going down the route of not doing it.

Rajeev: Absolutely, Srini. One great example is the landlines that if you think of are destroyed. There are no landlines. Another question from Eravi is: which industries — and perhaps dovetails into what you were mentioning — which industries are going to get destroyed? Which industries are going to emerge because of this transformation that we’re going through?

Srini: If I talk southbound for a minute, I think the entire telco vendor landscape will change. Just as telcos themselves have been monolithic. I think our vendors have been monolithic as well. So traditionally, for example, the way you buy a network is you buy a network in a box which comes with software and customized hardware. Now you only need to take a look at someone like Rakuten in Japan who’s tearing that model apart. Now what you have is kind of bare metal with software running on top of it. This is for stuff that only three years ago we said could never land up being disaggregated which is a mobile base station. A mobile base station today is getting disaggregated a bit like what happened years ago with cloud computing, where suddenly all of the large stuff you had in terms of racks of servers landed up getting decimated.

I think the telco layer vendor landscape will come under pressure. I think Telcos themselves, if they don’t softwarize, will come under pressure as well. There will be more pan-national across geography models, which will emerge, because things like software are a lot easier to port across borders than physical infrastructures.

Rajeev: Absolutely. Srini, there’s a question from Phil: How can Telcos scale down to smaller innovators and companies that have value to add?

Srini: It’s a great question. I can tell you what we’ve tried to do with mixed success. I really like things like — we have fun called DT Capital Partners, which is a venture fund, which helps bring in some of the fresh thinking from the smaller innovators into the telco. I think the reality is part of what makes a telco hard to deal with is also just processes and governance, apart from willingness to fail, and willingness to experiment.

So I think a big thing for Telcos for me is also not everything needs to start large. Specific pilots on new technology are often a lot more valuable than going large. With our infrastructure mindset, we tend to basically either go larger or go home, which is obviously an issue.

Rajeev: That’s right. That’s exactly right. You remind me of a conversation that I was having with a financial services executive a few months ago. They were trying to create a completely new business model in the landscape, and the immediate goal that everyone had was for that business to be a couple of billion dollars in revenue. Exactly to your point — getting that business to a few 10s of millions is an amazing achievement, given that they are changing the operating model. So, I completely understand what that means.

One other question around the industries that can take advantage of what Deustche Telekom is doing and what the telecom industry is expected to do: what do they need to do in terms of their culture, in terms of their mindset, in terms of things that they should be adopting, so that they can be more successful?

Srini: I think there’s a few big industries that could could take advantage of it. Most cloud native technologies that operate in the telco space, or in the near telco space, is obviously a big place. My advice to people working with Telcos would be a couple of things. One: within every telco there are always pockets of people who are forward leaning and who do want to bring in new technology. Often what matters more is how passionate your sponsors versus how senior they are, because it’s not necessarily common to find a combination of someone being senior and passionate about new technology. Making it succeed is often worth the patience, even though you have to put up with the non-digital processes of a telco.

Srini: I think a lot of startups I’ve seen trying to work with Telcos fail because they start with “You guys are so bureaucratic. You’re not like us.” That for me is kind of tautological. So what’s the insight there? The question is how do you land up, despite the greater bureaucracy and the processes of the telco, actually sticking through it and making a few pilots succeed, and then being able to scale that up. Being frustrated and trying to change the telco ain’t gonna work.

Rajeev: Absolutely. It’s about it’s about learning. It’s about learning from failure, and it’s about adapting to be successful. There’s a question on 5G, and 5G is pretty controversial. There are some people who support 5G. Some people think that it’s many years away. What are your thoughts on 5G?

Srini: I think 5G will go through phases of evolution. Let me start back with what I think is the most exciting part of 5G. The most exciting part of 5G without a question for me is the ability to slice networks. The 5G core is the first scale technology we have which will allow you to effectively create a network of one — which is the speed and latency you get can be different from the speed and latency someone sitting next to you gets, which therefore really creates the ability to handle multiple use cases at the same point. I think that’s at scale first going to be a BWB technology, and then a B2C technology. I also think we’re probably looking at four to five years before that matures at scale, because you need a new core implementation and slicing, etc, etc.

I think in the period from now to then there are a couple of other pieces of 5G which will help. I think 5G is just a better protocol. So much as this may sound boring — it will be better speeds and a better network experience, and therefore allow you to do more on the existing box standard telecom product. I also think what you will see is the emergence of things called campus networks where even before you have slicing, a combination of 5G and WiFi 6 could land up really making progress, especially in industry automation. I think last but not least, 5G and the protocol will allow us to at least continue producing gigabytes at a ever declining cost per gigabyte, which means you can continue to deal with greater data consumption. So, if I work back from slicing to just basic efficiency and economics, I think there’s a lot 5g has to offer. I think like most things, though, you need to be careful that you don’t see it only as slicing, which is four to five years off, and be disappointed with everything else in the meanwhile.

Rajeev: That’s absolutely right, Srini. Absolutely. It’s going to take many, many years for it to come to through your fruition. There’s a question from Deepak around COVID: What was the impact on Deutsche Telekom for COVID, and how did you all address that and overcome it?

Srini: I think COVID was a real challenge at one level, and this is much more at the physical infrastructure level, right? We had 80% of Deutscheland go into home working overnight. You had a similar phenomenon in the US, here, and across Europe. Where we run fixed networks, it was for obvious reasons challenging because of the amount of capacity that had to go in.

I think on the whole, while COVID’s obviously been hugely unfortunate in terms of the impact on the economy, it’s been not easy for our employees to deal with. We’ve had people going out there and repairing devices in homes in the middle of COVID. You can imagine the fear and the nervousness with all of that.

One thing it’s brought —I think maybe it’s a perception in society, that it’s brought the ability to something that’s an enabler that’s essential for modern life. We’ve seen that in brand score after brand score when we look at which industries have responded best to COVID or which players have responded best to COVID? I think we’re scoring, certainly in Europe, almost twice as high as the next industry or the next player. So there’s a bit of a silver lining, which is I think, as infrastructure providers how central and how system relevant we are has only been underscored. It’s been tough to deal with a bunch of it.

Rajeev: Absolutely, I can completely understand Srini. I mean, it’s an essential service, and how can we make it the right balance. We have a minute to go. There are lots of questions coming up. We’re going to ask you one last question: When we see your LinkedIn posts, we see a lot of your posts in German. How have you gone through learning German in such a short timeframe?

Srini: So firstly, my Germans is barely up to scratch. I can understand about 80% of a meeting in German, but I can probably speak about 10% of it. So, I’m still learning. I spend an hour every day — 7 to 8 every morning is learning German.

Rajeev: That’s incredible. That’s incredible, Srini. I think that’s a lesson for all of us in terms of what we’re passionate about. Srini, any other thoughts or any final comments?

Srini: I think the coming together of software and physical infrastructure is an incredibly exciting journey and one that I think will, if we play this right as an industry, enable us to be even more open to partnerships and even more open to working together to create a larger pie for everyone.

Rajeev: Super well said, Srini. Thank you so much for taking the time to speak.