Over the course of the last couple of months, I've had the opportunity to talk to senior leaders in major telecom companies in North America and Europe. In almost every conversation, I'm reminded of the myth of Sisyphus, the Greek king of Corinth, who, because of his deceit, was condemned by the gods to push a large rock up a steep hill, only to see it roll back down again when he got it to the top, and to repeat the same task endlessly till the end of time. Like Sisyphus, telecom operators seem condemned to an endless task of investing magnificent sums of money in bigger, faster pipes with denser capacity, to bring new services over their networks to their customers, only to see it all roll down the hill when Over-The-Top (OTT) providers swoop in with their apps and commoditize the telecom network to just bit pipes and take revenue directly from the end users bypassing the telecom provider entirely.
So what is OTT? Simply put, OTT is telecom parlance for any online service you use over the network of your internet service provider (ISP), whether cable, fixed line or wireless (on your phone or on your computer). Most of us already use OTT services like Netflix, Hulu, YouTube, Spotify, WhatsApp, Skype, iMessage, and Facetime without actually realizing it. OTT services built on cheap, elastic public clouds deliver voice, messaging, and streaming audio/video content directly to consumers over the internet, and are dominating the media and telecom markets today. In core telecommunication services, messaging and voice apps like WhatsApp, Viber, and Apple’s iMessage account for more than 80 percent of all messaging traffic on their networks, while Skype alone accounts for more than a third of all international voice traffic minutes (and is the world's largest long-distance voice provider). This pattern has led to steep declines in the average revenue per user; with minimal revenue growth; and ever-tightening margins for telecom providers in their core business.
With data, the story is even more frightening (especially if you are a telecom executive or shareholder). Almost every telecom operator, whether wireless or wireline, has been pushed to become just another bit pipe provider. Almost 70% of the data traffic of a telecom provider is video content coming from OTT services, like Netflix, YouTube, or social media. And it’s a perfect catch 22 situation for the telecom provider: either provide a fast bit pipe to keep your customers or see them go to a competitor willing to live on thinner margins. What makes OTT a particularly challenging contender to deal with is that it’s built on the public cloud and thus enjoys a very strong set of economic benefits in being able to scale up and down without the need to own any of the infrastructure needed to deliver the OTT service. OTTs are just one of the many manifestations of the disruption that the public cloud has created in telecom providers’ core business by enabling applications to be built cheaply and quickly with little to no dependence on the underlying infrastructure.
Edge to the rescue
To remain relevant in this hyper-disruption unleashed by the public cloud via OTTs, telecom providers need to identify new growth areas that combine the great potential of a brand new addressable market and telecom providers’ existing core competencies. Edge computing sits squarely at the intersection of these two criteria, and telecom providers who embrace it could possibly end up successfully surfing the tsunami unleashed by OTT apps and services on top of cheap public cloud computing.
I argued in my previous post that edge computing has the potential to eclipse what we call cloud computing today: a centralized architecture built on hyperscale data centers located in places with access to cheap and abundant power and, conversely, far away from most humans and devices and therefore unfit for emerging applications that need low latency, that process large data streams in real time closer to the data sources, or that have bi-directional traffic (instead of just downstream cloud-to-client traffic patterns). Edge computing, unlike the public cloud, will be built on either infrastructure owned by telecom providers or on new bespoke microdata centers with direct access to Radio Access Networks (RANS) (like Vapor IO's incredibly exciting vapor edge offering) in dense urban environments where most humans and their devices are within 50 ms of latency.
Turning the Edge into a strategic weapon against public clouds and OTT apps
To succeed with edge computing, telecom providers need to co-opt application developers away from public cloud platforms to using their edge platforms to write future applications and to pre-empt OTTs from commoditizing their next generation infrastructure.
These edge native applications can then be delivered as managed services by the telecom vendors themselves, creating new revenue streams and competitive moats to fend off the threat of public clouds and OTT apps and services.
Co-opting the developers away from public clouds will not be an easy problem and will require telecom providers to solve three key technology challenges, either by acquiring or partnering suitable software providers:
- Distributed data: The edge, unlike centralized hyper-scale data centers, is a geographically distributed and massively decentralized infrastructure (with no single locus of centralization) and at its core is a massively scalable decentralized and distributed database. Today's distributed database and storage systems are not sufficiently scalable to solve the data consistency and consensus when running across thousands of distributed nodes as a single logical programmable database (I plan to make a future post to elaborate on this aspect).
- Easily programmable: The edge must be programmable like a public cloud by offering a Container or Platform As A Service (CAAS/PAAS) abstractions for application developers to work with while hiding the complexities of the distributed architecture across apps running and storing states on tens/hundreds or even thousands of edge nodes. This is the central technology problem that will need to be unlocked to truly turn edge computing into a mainstream offering and a contender to take on the current model of centralized public clouds and OTT apps and services that run on them.
- Global reach: Telecom providers, by their nature, are regional and don't operate globally the way public cloud providers do (due in part to the legacy of deregulation and licensing of the spectrum). Therefore telecom providers will need to form one or more global alliances to provide one contiguous global service across networks for edge computing and to hide the complexities of disparate networks and infrastructures from end users. The good news is that the telecom providers already have deep expertise in this: global roaming, for example, is one way that telecom providers currently provide seamless voice and data connectivity across national boundaries for their subscribers.
Bringing together such an edge cloud architecture that tightly couples (1) a globally available database with (2) a rich CAAS/PAAS platform in (3) a global federation of edge clouds stitched together to provide data and code mobility would enable telecom providers to provide a true alternative to public clouds for app developers, but with the benefit of allowing applications to run in very close proximity to end users and devices—an advantage that hyperscale centralized public clouds will find very hard to compete against. With such a combination in place, telecom providers would serve as the critical platform for the complex data flow that links machine sensors to back-end services, in real-time analytics or cybersecurity, and the transportation supply-chain infrastructure to navigation, factory floors, social media, online multiplayer gaming, and autonomous consumer and business logistics apps (such as self-driving vehicles). Machine-to-machine revenues are projected to grow by double digits over the next few years, so they could potentially create massive new revenue streams in areas that are core to telecom providers and provide new greenfield opportunities to build and fortify activity away from public cloud providers and their OTT offerings.
But let me end this post by pointing out the obvious: the edge is too important for telecom providers to ignore. In the previous innings of the cloud game, everyone (including yours truly) underestimated the potential of the public cloud to change how we consume media and apps, and was taken by surprise at how quickly it went mainstream as an application development platform and created brand new billion dollar OTT businesses while rapidly commoditizing and disrupting so many incumbents (I will spare you the recanting of various tales, such as Blockbuster vs. Netflix or WhatsApp vs. SMS). Telecom providers cannot afford to make that mistake with the edge. It's now the proverbial bottom of the 9th innings, the bases are loaded, and the telecom providers are at the bat—it’s anyone's game to win.