In 2017, reality finally caught up with the hype. After trials, evaluations, and proof of concepts, network functions virtualization (NFV) is finally moving into deployment. Yet for many operators, the transition hasn’t been a walk in the park — or more aptly, in the clouds. Operators have faced challenges, both financial and technical, and have had to weather turbulence in their transition to the cloud.

Often when operators get serious with NFV and rubber hits the road, finance is the key driver. It is a directive that usually comes from an executive. Some operator teams who are at the coalface of NFV implementation have confided that funding is too drawn out, confusing, and incompatible with the goals of virtualization. This is especially true at the early stages of platform rollouts targeted for the cloud environment. Operators struggle to track capex and opex spend with traditional IT fulfillment. The irony for operators who aspire to migrate to the cloud is that more than 80-90 percent of their budget is earmarked as capex while a cloud model inherently demands a pure pay-per-use opex model.

Moving to the subject of return on investment (ROI), our experience suggests that operators are assessing their returns on specific virtual network functions (VNF) use cases. Given the phenomenal growth in mobile video, video traffic management has been a significant VNF use case. A large tier 1 Middle East group operator recently deployed a traffic management solution on a full NFV platform. The ROI was calculated on a video traffic management VNF that reduced video stalling by over 30 percent, while the virtualized infrastructure manager (VIM) was accounted as a pure cost item.

Therefore, from an investment perspective, we expect operators to spend over 70 percent on VNFs with the balance and 30 percent on NFV infrastructure. As a result of this, large network equipment providers (NEPs) are working with specialist software vendors to develop VNFs. In some cases, NEPs even provide funding for operators to deploy VIMs. Their goal is to recoup costs and make a profit over a five-year horizon, hoping that operators deploy more VNFs and secure new revenue streams.

As NEPs promote and push for network providers to take up their VIMs, operators must be careful to avoid vendor lock-in. This might seem contradictory given that NFV is all about building a truly homogeneous platform across all network silos. Not so. Some operators are ending up with fragmented NFV environments, the very problem that NFV had set out to solve.

Operators are also the victims of their own success. Subscribers have an insatiable appetite for the latest phones with the highest possible specs and fastest network speeds. To deliver this, operators have had no choice but to invest and outsource the development of NFV systems to large vendors. Owing to this, parts of the industry are witnessing NFV silos in vGiLAN, Virtualized IP Multimedia Subsystems (vIMS), etc. as each NEP comes with its own NFV manager that often operates in isolation. That is a shame given that NFV was all about open architectures and seamless interoperability.

Can an operator purchase infrastructure from multiple vendors to avoid this risk? Unfortunately, not. Current infrastructure available from NEPs has not reached sufficient maturity to make this a reality.

It is not all bad news – there’s a silver lining in that NFV cloud. As NEM vendors look to win contracts, operators should negotiate advantageous licensing arrangements. A business case that is based on hardware alone needs to take into account that today’s commercial off the shelf (COTS) environment is not as efficient as appliance hardware. Operators need to factor in hardware, licensing, and process changes into the business case.

The functions provided in the operator’s private cloud must have clear operational and functional key performance indicators (KPIs). This provides the data to not only demonstrate that a given service chain is up and running but crucially, that NFV is delivering the value intended for the operator. To enjoy the full benefits of NFV deployment, an operator needs to be agile. It needs to be able to move fast and take advantage of changing market conditions. And yes, it also needs be innovative.

Once in a while, you get an operator like T-Mobile in the US with a CEO like John Legere who tears up the traditional telco playbook and takes some bold risks. As Legere demonstrated with mobile video with his creative Binge On package, NFV similarly will require bold top management leadership to move forward. AT&T’s Chief Strategy Officer, John Donovan, has introduced a number of game-changing initiatives such as more widespread adoption of open source software and outlining clear objectives tied to annual bonus in order to drive NFV across his organization. This has helped prevent the silo traps and related challenges that we see mushrooming in other operators. This is not to say that everything has been smooth sailing at AT&T but then, is that a realistic expectation when one is truly an early adopter of a disruptive force like NFV?

Can operators fly stress-free to the cloud? Yes, but there have been a few bumps on the way. That’s to be expected given that NFV is still maturing. Operators need the right business and agile technology process along with effective leadership to make NFV a success. Otherwise, NFV ends up being an incremental and evolutionary change that’s hamstrung and not something that’s truly revolutionary.