Wednesday, October 24, 2007

Microsoft OCS 2007 Launch

I attended the launch event last week and I have been thinking over what I really wanted to say about OCS. First, I think it is very cool product (Kudos to the team). Second, I think Microsoft is not marketing it correctly in the Enterprise space. Third, they have definately gotten Cisco's attention (I think I ran into more Cisco folks I know at the event then Microsoft folks if you can believe that.) Fourth, the guitar opening for the keynote was lame.

Ok, so, to my points. If you haven't played with OCS 2007 you should. I think the new MOC and presence features that they have put into this platform are very cool. If you are looking for a great presence and collaboration tool this should be on your list without a question.

Microsoft has been delivering a message of VoIP as you are (which just makes me think of a Nirvana song) and the majority of their published presentations on the topic of QoS is to say it is more important to deal with Quality of Experience (QoE.) They have taken the approach of Skype and Live Meeting and are utilizing a wideband codec that is very good at delivering high fidelity audio over questionable networks (think the unknown of the Internet.) Their argument is that if you design for the potential loss and latency you commonly see on the Internet then the product will perform just fine in a corporate enterprise environment. They have tailored their design philosophy this way and therefore they arrive at the QoE argument. They are not concerned with what the network is doing with QoS, they argue that the advances in the wideband codec handle those issues in almost all cases. They do request that you add bandwidth but they are saying do not bother to design QoS into the network at all.

Well, here is where I think they are missing the mark. Most enterprises UNDER provision in networks - especially on WAN's. No one ever gets praised for increasing a companies monthly recurring costs so for many enterprises adding more bandwidth is a very expensive proposition. In addition, many utilize QoS to prioritize line of business applications, data backup, replication, database functions, citrix and lots of other network traffic due to the limitations. They are willing to keeps costs in check and build a more measured approach to controlling their costs relative to what all their application needs are. With voice and video, they are simply another application riding on top of IP and therefore should be designed and accounted for in the same way. I think Microsoft is missing the main point for the enterprise, they need to redo what they are telling partners and customers about the best strategy for deploying because they have not accounted for the fact that for many companies they don't have a WAN that has unlimited BW potential. Perhaps they could argue that no one should get a WAN and that all infrastructure should run on the Internet since getting a relatively large pipe to the Internet is cheap - there may be something to that but we don't have room to talk about that now.

Since I work for a company that is both a Cisco Premier Partner and a Microsoft Certified Partner (plus others) it is interesting seeing the shake up at Cisco about what Microsoft is doing. I think Cisco is in an interesting position of having a much more mature and stable platform but one that is modeled in many ways on a traditional voice PBX. Microsoft has really stepped out of that model and is shaking things up. They are a long way away from having the feature sets and robustness that Cisco has today but some of the features that Microsoft has out in a 1.0 product is very impressive.

Well, more thoughts later but that gets the first wave done.
- Ed