Changing Mindset to Create User Experience – Driven Operation Model


Author: Eric Xu, Huawei Rotating and Acting CEO

Telecom has been a lively industry over the past three years. We’ve heard many new terms, and indulged in intense discussions. Software-defined, virtualization, SDN, NFV, cloud data center, 4.5G, 5G, and ultra-broadband are just a few examples. But it’s a pity that the operation was not as in fashion as the network. Why? Fundamentally I think it is a matter of mindset. We’ve become used to a network-centric approach, always looking at the network before operation, which is considered as a support function for the network. The environment however, is far more different, especially with the disruptive operation model of Internet service providers. It is relevant to draw upon experience of the Internet and revisit the operation model of the telecom industry. I want to offer my thoughts on three topics, the new perspective on operation; the different way of thinking for operation transformation and IT restructuring; and the relationship between infrastructure evolution and operation transformation. In a nutshell, I believe a mindset shift is needed for the telecom industry to create a new operation model based on user experience.

The first observation is the new perspective on operation. We need to start from the end users, and then decide what kind of operation model would be needed to deliver the right user experience.

The term user experience has two elements, experience in using the services and experience in obtaining the services. When users make phone calls, text messages, or browse the web, it’s part of the service usage experience. But the user experience is much more. It also includes users’ experience in obtaining the services. How to find and buy the services? How the transaction is done? How is after-sales service? The answers to these questions sometimes are more important than the service itself. Clumsy transaction and lengthy provisioning may scare users away from the services. To me this might be our single biggest gap compared to OTT players.

Then how will the future experience look like in obtaining the services? We’ve come up with a word to summarize it, ROADS, which stands for Real-time, On-demand, All online, DIY, and Social. ROADS well captures the consumption behavior of a generation that has grown up with the Internet over the past twenty years. They are the digital natives that are shaping the future; they ARE the future as a matter of fact.

If we look at the telecom operation model from the perspective of consumer behavior and desired user experience, we’ll find that a major transition is required for the IT system (OSS and BSS), from a mere internal support system to a production system that turns telecom operators’ networks, services, content, and other assets into compelling product and service offering that users can easily access and utilize. More importantly, it’s much more than simply changing the IT system; it is in essence an operation transformation, one that needs to start with user experience and harmonize organization, processes, IT, and infrastructure.

My second thought is about the different way of thinking for operation transformation and IT restructuring. The design principles, operation models, and technology architecture applied by the Internet companies need to be leveraged to shape the operation transformation of the telecom industry.

First, the focus of operation needs to evolve from user experience of the network to user experience of services. User experience today is primarily defined and measured through network KPIs, like bandwidth, latency, call drop rate or call completion rate. This is right, but not enough. Operators need to go further to organize their operation around service experience KPIs. And it’s not just about services, but also how services are found, purchased, and used, and how after-sales services are offered. It has to be an end-to-end experience. At the end of the day, network metrics are not perceivable to users. What’s relevant are the metrics that measure the services they use, like voice, video, leased line, VPN, or cloud services, and the metrics that measure the experience how these services are used, like service provisioning in minutes.

Second, the operation model will need to be built on real-time, autonomous systems, rather than manual systems. Large Internet companies employ much smaller operation teams to serve a significant number of data centers, servers, and users; they are much more efficient compared to telecom companies. The key to efficiency is the underlying design principle. The telecom mindset is maintenance oriented – telecom operators only feel assured with human intervention and control. Being real-time and autonomous is the Internet approach. Services are easily scalable and provisioned in minutes. Faults get isolated and self-heal. The Internet systems look quite simple, but they are technically more complex. As such, there has to be a fundamental shift in the telecom operation model, and systems need to become autonomous. To do that, network autonomy is needed with new IT systems at the front end and new network control systems at the back end; traditional EMS and OSS should be simplified into monitoring and alarm systems; and service provisioning, scaling, fault isolation and recovery have to be automated.

Third, IT systems will move from a closed architecture to a cloud-based Internet architecture. With a new positioning for telecom IT systems, from internal support to external service, technical architecture will be fundamentally different. There might be only thousands of internal users, but the number of external users is in the order of hundreds of millions. In light of these new requirements, traditional scale-up architecture is insufficient in many respects, for example, openness, ease of use, and scalability. A scale-out, cloud-based Internet architecture is the only way to support a massive number of users. This is a natural development driven by the changing user needs and the advancement of technology.

Therefore operation transformation and IT restructuring need to be shaped with a different and more advanced design philosophy. With the new approach, system and technology architecture, as well as the ultimate outcome, will be very different.

The third observation is about the relationship between infrastructure evolution and operation transformation. They are highly interdependent. Software-defined infrastructure of the future will set the stage for operation transformation, but also present us with new challenges to cope with.

Over the past ten years, network architecture has been evolving toward ALL IP. Over the next decade, virtualized, cloud, and ultimately software-defined architecture will take shape. This is a broad consensus across the industry after years of discussions. It’s also the foundation for operation transformation. Many problems may be perceived as operational issues from consumer behavior and user experience point of view, but the fundamental challenge lies with infrastructure. Today telecom operators cannot provide on-line services on demand and in real time. The bottleneck is not the operation system, but infrastructure; infrastructure is not automated and intelligent as it needs to be. Therefore, while new developments that we see today, like SDN, NFV, and cloud IT infrastructure, present telecom operation system with new requirements, they are also the key building blocks for automated and intelligent operation. For operators, without automated, cloud, and scalable infrastructure, a ROADS experience cannot be delivered.

With software-defined architecture, cloud data centers are becoming the foundation of the telecom network. It’s also changing the way networks are deployed, from node by node, to layer by layer. These developments present unprecedented challenges to the ecosystem and the operation model. Through software-defined development, network equipment will no longer be hardware/software integrated, and cloud data centers will become the foundation of the network. As a result, network deployment is implemented layer by layer, not node by node. This is a big change for the industry. The first impact is on the industry chain. It’s no easy task that products in different domains and from different vendors can come together to form a system that can be delivered and beautifully operated. The industry chain will also develop differently; it will be built upon the ecosystem, and no longer driven by standard bodies. The second impact is on operation. As the network moves to the cloud, SDN, and NFV, maintenance of hardware infrastructure will be separated from service deployment and provisioning, as everything will be software-defined and cloud-based. Hardware/software decoupling will make it very difficult for issue demarcation and location. In particular, network complexity is growing as software problems are sometimes caused by hardware. As a result, orchestration between hardware and software is needed. These are the changes that the operation system will have to adapt to.

Therefore, evolution of the infrastructure and transformation of the operation system must go hand in hand, and be consistently planned and implemented around user experience.

All these changes and developments are easier said than done; many issues still need to be addressed and resolved. How to define KPIs for the user experience? How to design target architecture for the future? How to build unified IT infrastructure? How to move forward from today into the future? What strategy and pace for the evolution? How to better collaborate between operators and suppliers? And the most difficult of all, how to free ourselves from our decades-old mindset and establish the right corporate culture and organization to support the future model? To answer these questions, a mindset shift will be necessary, and user experience has to be at the core to shape the way we think and act.


Leave a comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.