AT&T, vCPE and Vendor Differentiation – Can’t Say We Didn’t See This Coming

Peter Jarich
Peter Jarich

Summary Bullets:

  • AT&T’s recent “Network on Demand” announcements with Brocade, Cisco, and Juniper point to its continuing progress with virtualization and its Domain 2.0 strategy.
  • The fact that three vendors are supporting the launch and supporting virtual routing on a common platform suggests NFV is beginning to deliver on the promise of vendor agnosticism.
  • It also suggests that vendor messaging needs to evolve and focus on VNF performance vs. availability.

Glen Hunt from our transport and routing team recently wrote an analysis of AT&T’s “Network on Demand” work with Brocade, Cisco and Juniper. It’s a good report. Among other things, it calls out the significance of AT&T delivering virtual network functions (VNFs) from different vendors, out to the enterprise, as a solid proof point of NFV’s vendor agnosticism promised. At the same time, Glen also points out that the first “Universal CPE” being used in the service deployment was custom-built to AT&T specifications by Juniper. On paper, at least, the concepts of “vendor-agnostic” and “custom-built” are contradictory.

In other words, pioneering operators like AT&T are making progress with NFV, but we’re still a ways from reaching the “promised land,” potentially because it’s not in anyone’s interest to be too agnostic. Implied in all of this is another critical takeaway: vendor messaging around NFV needs to evolve.

To date, marketing around VNFs hasn’t needed to be about much more than availability. Does a vendor offer a specific VNF? Has it been put into commercial service? Trials or PoCs? What you have at AT&T, now, is three different virtual routing functions (as an example) being supported on a common platform to deliver services out to the enterprise. How will a given enterprise choose between them? The same way they’ve always made these decisions; by factoring in price, performance, brand affinity, solution engineering, etc. Map this onto a service provider’s procurement process and you see the requirement. VNF sales and marketing efforts are about to get a lot more complex.

For an industry still largely in its infancy, this might sound daunting (at least if you hope to get it right). Luckily, there’s some good news.

  • We’ve All Seen this Coming. Nobody was under the impression that NFV marketing wouldn’t need to get more sophisticated and competitive as the market evolved – at least I hope not. In our VNF catalog report from last month, we even called this out as a necessity as NFV becomes the “New Normal.”
  • Back to the Future. What will vendor VNF marketing look like going forward? If it focuses on price, performance, and broader solution packaging (which it should) then it will look a lot like their marketing has traditionally. Sure, it will need to focus on software more than hardware, but the basics remain the same.
  • Cisco as a Case Study. It’s unfair to say that vendors are solely marketing VNFs based on availability. Larger solution capabilities do come into play, for example. Cisco, however, takes this one step further by leading with a service story and selling NFV-enabled services instead of NFV network solutions. Perhaps this is why it perennially comes out at the top of our service provider surveys of NFV vendor perception.

Leave a Reply