Deutsche Telekom’s CIO on how new approaches to IT can drive growth

In October 2024 Deutsche Telekom CEO, Tim Höttges, surprised the telecoms industry when he said that investment in 5G and fiber networks was delivering financial returns. Speaking during the company’s Capital Markets Day, a principal reason he gave for the success is that high-quality networks attract more customers.

He also said he expects the company’s profitability to increase by more than 5% due to “the added value you can generate.” Here the onus is on IT: “What we need for that is the IT capabilities, which we have.”

Inform spoke to the executive in charge of Deutsche Telekom’s IT engine, CIO Peter Leukert, at his offices in Bonn to find out what his teams are doing to support future business growth amid rapid 5G and fiber network rollouts.

In Germany alone, the Group expects to pass around 2.5 million new homes each year with fiber by 2027, taking the total number of homes passed to around 17.5 million.

“We need to support the scale and productivity of the [network] connectivity provision at a pace and a scale which hasn’t been there before,” explains Leukert in a video interview with Inform.

At the same time there is a requirement to continually enhance customer experience because “the benchmark for the customer experience is … actually the OTT player,” he adds.

It should therefore come as no surprise that Leukert emphasizes the importance of rapid time to market.

Adapting to market change

“You can't predict what the customer needs and how the customers’ needs will change. So, you need to be agile. You need to be fast,” he says. “For me, time to market is one of the key KPIs that shows how much value we can deliver to the business.”

Faster time to market, however, requires IT teams to be flexible and able to scale – and they must collaborate closely with the business: Leukert, for example, exposes developers to customer feedback. “Culturally, we make our teams customer obsessed.”

One example of how the IT team’s speed and flexibility works in practice is its development of a new operational support system [OSS] to support the company’s fiber roll out.

The company initially sought an off-the-shelf system. However, it realized it would be faster and more efficient to develop an in-house OSS that was tailored to the regulatory constraints and service requirements of the German fiber market, while integrating with Deutsche Telekom’s existing IT systems.

“We designed and built from scratch a new OSS system for fiber in nine months,” explains Leukert. The team has been able to expand the scope of the system to address different kinds of fiber services, such as to the home, or multi-dwelling units, as well as subsidized and non-subsidized services. It has also attracted external interest: Two of Deutsche Telekom’s partners now use the OSS.

Measuring business value

Several other business metrics point to the IT team’s success.

“The number one is scale: We can handle a factor of 10, 15 or 20 times increase in simultaneous real-time customer interactions during [service demand] peaks such as soccer games.”

Another key measure is “the revenue, time-to-market and ARPU that you can defend as a premium provider.”

A third gauge is employee productivity. A combination of better digital channels and IT support for customer agents has delivered a 50% increase in productivity in Deutsche Telekom’s customer services organization, according to Leukert, while reducing customer complaints by a factor of five.

Multiplying productivity

GenAI promises further productivity increases.

“GenAI will, we believe, enable another 50% increase in productivity in customer service, and increase customer satisfaction.”

For example, it has enabled the company’s consumer chatbot ‘Ask Magenta’ “to crack cases that the traditional chatbot couldn’t,” with “a very, very high level of predictability.”

GenAI’s impact extends to Leukert’s own teams. “It will make a big difference to how we build code; how we fix vulnerabilities in code; how we do predictive maintenance in operations.”

Nonetheless, successful AI deployment still depends on mastering certain IT fundamentals.

“Some things never change. If you really want to really have an impact with AI, the [key] is not building or training the models. The key is integration” says Leukert.

And when Ask Magenta accurately resets a customer password using GenAI it can thank careful integration with backend systems.

“The importance of integration and understanding your own data and business content [is true] with every new technology, but in particular with AI models,” points out Leukert. “You have to be careful … to show you are taking decisions ethically and don’t have bad biases in the model.”

For this to happen, a telco needs “good architecture, good governance and model-agnostic ways to integrate [AI] features, explains Leukert. It also requires a sound data strategy: “AI reminds us of the importance of clean data and meta-data.”

The network effect

It’s not only AI that is turning the spotlight on data. As networks become more software-based, and the industry gears up to deliver future intent-based services, telcos will need to combine data in new ways.

“We are still at the beginning of unlocking the full value of data, particularly when bringing together the real-time or near-time technical data from the network with customer experience data and going to one-person micro-personalization,” says Leukert.

He believes that the distributed nature of networks means it will not be “a one size fits all pattern” based on a central data lake.

Instead, tasks such as the adjustment of customer profiles or applying energy saving across mobile access networks, will rely on decentralized data.

“There will be a decentralized element to the data strategy in the network area [and] … then the most important results and insights will then need to link back to your central data platforms,” he adds.

In the meantime, the development of software-based networks is fostering collaboration between Deutsche Telekom’s IT and network teams.

”We use the same continuous integration, continuous deployment [CI/CD] chain across IT and the network area and across several of the operating countries in DT. Another common enabler is API gateways,” he says, adding: “I think that's a that's a big achievement.”

The blurring of the boundaries between the OSS layer and network functions is another factor bringing IT and network teams together.

An example of joint innovation includes the building of an Open RAN management system. “People from the network area understand latency… IT understands the decentralized data,” says Leukert.

A changing supplier landscape

Deutsche Telekom is also changing how it works with its technology suppliers and partners.

“On the one hand, the collaboration with [and] the dependence on hyperscalers is increasing, along with the benefits you're getting from that,” details Leukert. “On the other hand, the dependency on some of the more traditional software vendors in the BSS area is decreasing as things are more decoupled and we build and use more open source.”

TM Forum’s Open Digital Architecture [ODA] is helping Deutsche Telekom innovate in how it sources and uses software.

“ODA is really cool, because it is leveling the playing field for those smaller [companies that] … don't need to do a huge investment in integration for each of the telcos they want to work with.

ODA “has also been incredibly successful in building software that we use across all our eleven operating countries in Europe,” according to Leukert.

“We have now a lot of software that we only build once, and one source code that we use in eleven countries. And you wouldn't get there if you didn't have a guideline like the open digital architecture that harmonizes across the different countries.”

Watch the video interview below: