Day 1 of Enterprise Connect: Cloud Confusion
Considering the relative immaturity of cloud UC, it might be too early to have definitive answers, but several good ideas came up.
My first day here at Enterprise Connect was highlighted by moderating a panel called "Choreographing the Cloud," which was intended to help the audience find the right mix of on premises and cloud based communications. The session had strong panelists with representatives from Alcatel-Lucent, IBM, Verizon Business, plus consultant Gary Audin of Delphi Inc., and it was one of the livelier panels I've moderated in a while.
Despite the liveliness, I don't think the panel fully answered the question of where companies should use public cloud, private cloud and premise-based solutions. Considering the relative immaturity of cloud UC, it might be too early to have definitive answers to that question. I do think though that it highlighted a number of issues as to why cloud hasn't been more fully adopted.
One of the first areas the panel probed on was the cloud value proposition. This was the most glaring hole in the cloud story. I asked each of the panelists why someone should consider cloud now and, across the board, I got answers that had to do with cost. True, it's lower cost and it becomes opex instead of capex and, in theory, should be cheaper to run than a premises solution. Gary Audin did point out that if you plan to keep the premises solution more than three years, it actually becomes cheaper to use a premises-based solution, which, in some ways, debunks the cloud value proposition.
Whether it's cheaper or not, I really don't like it when the focus turns to cost. If companies want to save money on telephony, call up your telco and hammer them on long distance. That's the easy way of knocking 10% off your bill. The focus of the cloud value proposition should be about doing things you can't do with premise-based communications. The most obvious answer here, and I was disappointed none of the panelists brought this up, is that cloud communications is a better solution for consumerization or "bring your own device".
Enabling a multi-OS, cross platform strategy is difficult, if not impossible with premise based solutions. In fact, this is one of the fundamental selling points of the Cisco HCS-based Sprint Complete Collaboration, which was announced earlier today at Enterprise Connect. I understand the solution providers want to use cost as a way to get customers interested, but if that's the main value proposition, the price will go to zero before anyone can monetize the investment. And that would be bad for everyone.
Another strong point raised, this time by someone in the audience, was around the interoperability and migration issues associated with a hybrid premise/cloud deployment. These issues would be the same for a multi-cloud solution as well. As everyone associated with this industry knows, interoperability between vendors is poor at best. When I talk about interoperability I mean more than just making calls and sharing presence info (although this can often be a challenge). Interoperability challenges such as creating a common dial plan, provisioning of users, change management, voice mail, etc. are all significant challenges in multi-vendor, hybrid environments.
Interoperability though could become the "killer app" for cloud services. If a cloud provider could provide full UC interoperability in the network, that would be a significant differentiator for cloud over premises. The cloud would become, in effect, the "Rosetta Stone" for UC. We're obviously a few years away from this, but it is a good stretch goal for the service providers. There are companies that do this for a subset of UC functionality such as GlowPoint and VidTel for video, but there isn't a cloud provider that does this for UC. I admit the ever-changing definition of UC might make this a utopian dream, but nailing voice, video and chat would be a good start.
The other notable topic of discussion was how to get started. There's obviously many ways an organization could get started and I felt fellow NoJitter colleague, Gary Audin, gave the best answer. His advice was to look at what UC capabilities the organization was missing and utilize those services from the cloud to shorten the time to deployment. He offered another suggestion as well, that a company could look at what skill sets they didn't have in house and use the cloud for that. So, cloud in this case fills an application or a skill set void. Of all the things I heard in the session, this seemed like the most actionable, practical advice for people looking at cloud UC today.