Plenty has been written about digital transformation, and it is commonly acknowledged that the biggest challenge when it comes to adopting new technologies is not technological. If new technology accounts for 20% of the deployment challenge, the other 80% revolves around how we change processes and cultures. The ‘traditional’ processes and culture that exist within telcos are not suited to agile deployment and operations.Modern innovation is embodied by the likes of Uber. It has no legacy, it was born on the cloud and grew up with DevOps. Uber used to claim that it enhanced its customer user experience every two weeks. Today’s average telco would be doing well if it changed its subscribers’ user experience more than once a year.
Telcos can’t start-up all over again, though. They can’t completely overhaul their processes and culture in one fell swoop. But they can change over time in small iterations.
Policies shape culture
Any organisation that is conditioned to guarantee certain levels of service or meet regulatory requirements will have institutionalised processes by having transformed them into policy. Telcos are almost entirely process driven, meaning they are almost entirely policy driven. These policies define the telco culture.
Telcos crave agility. They want to offer new services quickly, and they are starting to branch out into new markets. Take, for example, serving small and medium enterprises with IoT solutions. They’re using new technologies to do this, but they’re using their same traditional practices to deploy the new technologies. They’re delivering five-nines availability, because that’s policy. For some IoT services, such as self-driving cars, you will absolutely need five-nines reliability. But for others such as in-street advertising, you don’t.
Of course, some things will be harder for telcos to change. Networks require a lot of physical infrastructure including cell towers, base stations and fibre plant. These big infrastructure projects take time and they’re not going away. A lot of larger capital projects for telcos provide very little return at the start of their life, but towards the end of their duty-cycle they’re extremely profitable. This ROI model feeds into a culture that supports long duty-cycles. Telco fiscal policy says they can’t afford to fail fast.
When it comes to innovative new services, telcos need business cases that demonstrate ROI after a certain number of months, and they need approval from procurement to spend CapEx / OpEx. The dependency therefore becomes one where procurement has the control and the power over investing with little return - at least initially. This model does not encourage innovation.
Hybrid technology landscapes create hybrid cultures
When electing to move consumer and business communications services into the cloud, service providers are effectively decomposing capabilities that have historically been vertically integrated into network appliances (routers, switches, load balancers, session border controllers, etc) and recomposing them into a “cloud” made up of virtualization software managing off-the-shelf hardware. These recomposed services must be specifically architected to deliver SLAs on an infrastructure platform that by itself can only deliver 9.995 availability, must be integrated into existing operations environments, and must be continuously tested for its entire useful life.
Often current NFV-based services coexist as islands of capability inside a broader operations environment managing the “rest of the network.” New cloud management practices and processes (CI/CD, DevOps, configuration-as-code, etc), which are typically associated with agile development methodologies, must be harmonized with existing management practices (ITIL, eTOM, FCAPs, ITSM, etc), which are typically associated with waterfall development methodologies. Since agile and waterfall practices tend to be toxic to each other, the result is necessarily a hybrid of both.
Practical steps to affect cultural change
For all the reasons outlined above it would be nearly impossible for a telco to attempt a wholesale cultural change. Harmonizing operating models and aligning ecosystems and communities needs to be achieved in small increments over time by empowering teams and allowing organisations to evolve naturally.
When virtual network function (VNF) suppliers are allowed to dictate architectures for the supporting NFVI, VIM, MANO, SDN and automation tooling, the result is a vertically integrated solution that provides little or no economic advantages over the previous appliance-based approach. Telcos need to insist that VNF suppliers comply with their own platform blueprints based on open standards. They should mandate that the suppliers assist in integrating and operationalizing their VNFs into the operating environment. More importantly, the suppliers need to embed their experts into the telcos’ service deployment teams and hold them accountable to the commercial success metrics.
Telcos should also implement dedicated interoperability and regression testing environments. The combined contributions from different suppliers will almost always be unique to a given telco environment. Use of open standards and APIs will dramatically simplify interoperability, but it does not replace the need for high quality testing. Automating the testing can dramatically reduce time to market for new services.
When it comes to rethinking their organisations, telcos need to start small, iterate often and allow for failures. New approaches taken need to increase collaboration and communities to empower and inspire the organisation. If telcos feel the need to do something physical, they should put their operations and engineering teams in the same room, give them a goal or common mission, and ask them to come up with a solution to achieve it. Once that’s done, the service providers should disperse that group and create fresh cross functional teams for new projects. It will be the people who effect change, not the technology.
Visit Red Hat to learn more
www.redhat.com 
