In our recent blog posts, we have looked at different aspects of Total Cost of Ownership (TCO). CapEx spending gets the headlines and for good reason. Flattening revenues driven by lower revenue per bit coupled with explosive growth in bandwidth demands have made it difficult for carriers to achieve the industry’s standard goal of 15%–20% capital intensity, deﬁned as (the ratio of capital expenditure to revenue) at a time when there are more demands for that scarce capital budget. Service providers try to balance the competing demands by focusing on capital efficiency – how well spending generates profits. An analysis by Bain and Company describes the problem:
“Unfortunately, too many executives approach their capital intensity challenge solely through a lens of efficiency by trying to squeeze as many projects as possi¬ble into a restrictive capital envelope. The budget cuts and funding deferrals that follow draw the focus away from a fundamental truth: You cannot cost cut your way to long-term profit growth; you need to take market share. NSPs that fail to realize this head down a path that allows the competition to win.”
This argues that growing market share is the best way to optimize the returns on capital. The question becomes market share of what? It is critical for service providers to build share in the highest growth markets, following the adage of running to where the ball is going.
Of course, it’s not always simple. When we look at business services which have been a key profit driver for service providers, SD-WAN is growing very rapidly while MPLS service revenues have flattened. Vertical Systems pegged the US SD-WAN services market at $282 million in 2018. However, MPLS services are still huge with around $40 billion in worldwide revenues. There are lots of ways to characterize and analyze markets. Boston Consulting group’s product portfolio matrix (BCG matrix ) breaks sets up quadrants based on growth and market share and how they generate cash. In our business services example, SD-WAN is a rising star while MPLS is a cash cow.
Service providers are voting with their budgets on projects like cross haul and DCSG as the foundation for services like network slicing to MEC. We have discussed how these decisions are driving a massive expansion of routing as it gets pushed out all the way to the cell site.
There is one more important consideration and that is time to revenue from capital investments. For new services, you need to build out enough infrastructure to deliver the service. In a previous post, we talked about how legacy routers require an upfront investment and then capacity gets used up and added in a stairstep function. This will be more pronounced with new services. Service providers need to think about this lag from investment to ROI. Service velocity is about minimizing that time gap. Volta embraced a “pay as you grow” subscription pricing model to address this. You only pay for the software that generates revenue. Given the demands on capital resources, service providers need to think in terms of time to ROI.
Service agility is another related consideration. This is about how fast a service can be turned up or modified for a given customer. The faster this can be accomplished the sooner you can start sending out a bill and get paid. Agility is closely tied to OpEx because routing has relied so heavily on experts configuring with CLI tools. It is clear that the number of routers will increase dramatically, by as much as two orders of magnitude. While they will be smaller and much less expensive than legacy routers, the software that service providers need to drive services will be much the same. It is completely impractical to think that there are enough people to continue with manual processes with this kind of scale. It would be cost-prohibitive even if there were enough trained personnel.
Automation will be the key to leveraging the talent you already have to deliver this scale of services. For example, Volta has focused on using YANG service models supported by robust APIs that can interoperate with carrier-grade management and orchestration (MANO) platforms. There will be a transition, but service providers need to embrace platforms that can support this rather than perpetuating the old models.