Business process redesign is essential to NFV/SDN success

Michael Kennedy is a regular FierceTelecom columnist and is Principal Analyst at ACG Research

Michael Kennedy

Leading telecom operators began the network function virtualization (NFV) initiative several years ago in recognition of the need to reduce cost increases and increase revenue growth rates to sustain profitability. While much technical process on NFV and related SDN technology has been made, I believe business process redesign is needed to change the organization and people's jobs if the financial benefits are to be realized.

The industry seems to be well on its way to creating technical solutions that replace traditional purpose-built hardware with new approaches adapted from pioneering work done in the data center including virtualization, SDN and orchestration. Well-conceived solutions have been developed for virtualized CPE, cloud-based cable guides, cloud-based digital video recorder, transport SDN (multi-layer SDN), network analytics, virtualized routers and virtualized mobile core, among others. A few of these solutions have progressed beyond the prototyping stage and are now being implemented in production networks.

I have found that NFV/SDN solutions make their greatest financial impacts by reducing the labor content of many business-processes (40 percent to 70 percent decrease) and reducing process times by a factor of six. The shorter process times produce major financial benefits by eliminating much of the spare capacity assigned to maintain SLAs in today's networks and enabling service creation and delivery processes that rapidly deliver targeted services that fit customers' willingness to pay (value-pricing).

Virtual CPE and multilayer SDN provide two examples of the financial benefits of the NFV/SDN approach. Today, network operators deploy integrated service routers (ISR) at small and midsize business sites as the delivery vehicle for IP-VPN, managed security and managed router services. The ISRs cost the operators several thousand dollars on average and many truck rolls are required to install, service and upgrade services. The virtual CPE alternative employs a very simple onsite interface that rarely requires a truck roll while network and security services are provided automatically by virtual network functions in a cloud data center. The virtual solution reduces operating costs by at least 65 percent by eliminating most field service work as well as many of the manual tasks performed at the network operations center. Also much of the infrastructure work is taken from the network operations staff and turned over to data center personnel.

Today, most large networks are planned and built by separate transport and packet organizations. Manual information handoffs are made to coordinate the overall network plan and operation. Overall network capacity utilization is well below 50 percent due to the lack of timely information, uncoordinated plans and safety margins built into each organization's capacity estimates. Multi-layer SDN solutions dramatically reduce this built-in spare capacity by automating information exchanges as well as the planning and operation of the network itself. Many network planning and network operations jobs are eliminated. Also, as less construction is required many construction jobs are eliminated because labor is the largest component of construction work.

The NFV/SDN approach provides a powerful tool for reducing cost and increasing revenue because it attacks the primary cause of network operators' financial problem: manual work processes. This presents a dilemma that network and IT technology experts who have the responsibility, authority and expertise for NFV/SDN technology cannot resolve. Successful implementation of NFV/SDN, furthermore, will be personally painful in that workforce reductions will fall heavily on the network organization itself.

Organizational and staffing issues that I have encountered during my ongoing NFV/SDN work include turf issues in integrating the transport and packet organizations, resistance by the networking organization to using IT department data centers to host network functions, union resistance to cutbacks in the field service organization, failure to establish a compensation scheme to motivate product sales persons to sell virtual solutions, failure of vendor marketing teams to find an effective pricing scheme for software-based solutions, and failure of NFV/SDN technology architects to engage their network operations counterparts in the planning and development process.

Business process redesign projects such as those that were successfully implemented by the financial services industry in the 1990s are needed to achieve the results envisioned for the NFV/SDN initiative. Redesign must necessarily be driven top down and address the issues of key stakeholders including owners, customers and employees. The redesign effort begins by redesigning workflow (the process) to reduce costs and speed up the process. (The virtual CPE and multilayer SDN examples address this.) Next, technology is designed and procured to support the redesigned processes. Finally, the organization is redefined to support the new processes and technology. This involves defining new roles, required skills, compensation and reporting structure across the entire organization.

Cost reductions on the order of 50 percent are not obtained painlessly. The redesign of the financial services industry in the 1990s began with about 20,000 banks in the U.S. and ended with fewer than 5,000 banks. Along the way thousands of people lost their jobs. In some cases they were replaced with new people who possessed new skills required by the redesigned firms. In other cases entire firms failed or were picked apart by leveraged buyout specialists because they did not move fast enough.

My view is that the telecom industry does not yet find itself in the desperate position of the 1990s financial services firms and therefore redesign needed to capture the benefits of NFV/SDN will move slowly.

Michael Kennedy is a regular FierceTelecom columnist and is Principal Analyst at ACG Research, www.acgcc.com. He can be reached at [email protected]

Read more on