In the old days, the reluctance had a lot to do with the limitations, high cost, and relatively low functionality of Centrex. When VoIP came along, the Centrex carriers -- i.e., the local Bells -- never really put together compelling equivalents to what the enterprise could buy from the IP-PBX vendors, and IP-PBXs replaced Centrex in waves.
There's been a persistent idea that hosted IP telephony could be an appealing service, and wouldn't have to come from a telco. In theory, many types of players could buy a large-scale IP-PBX or carrier-grade softswitch like Broadsoft or Sylantro, at a fraction of the cost of the Class 5 switches that supported Centrex. Early on, there were predictions that systems integrators like IBM or EDS (pre-HP) might be the hosted IP telephony service providers of the future. But this market never materialized, either.
Not only did these services not take off, they weren't ever part of the conversation. But SaaS, or cloud computing if you prefer, are going to be part of the conversation now, thanks to Cisco.
The WebEx Connect offering hits all the right notes in the UC/collaboration symphony: Mashups, APIs, communications-enabled business processes. These are the essence of what people think will be valuable in communications technology that reaches beyond simply re-creating PBX features on an IP network. And embedding these capabilities inside the network opens up the potential for making them ubiquitous across an enterprise -- not in the sense that they'd roll out to everyone, but in that they'd be deployable to anyone, in a targeted fashion across offices and geographies, according to user requirements.
That would be appealing at any time in history, since it's still questionable that UC would be rolled out to every desktop everywhere; there are just too many people who don't need it at all, and too much variation in the kinds of work that people do, and thus in the kind of UC capabilities they need.
But it's especially appealing, as Sulkin points out, at a time when capital expenditures could be seriously constrained. If you can put communications capabilities in opex, you can move forward incrementally, hopefully through all but the worst of times.
The catch, as I wrote on No Jitter today, is that the resistance to network-based services is not going to just melt away in the face of the new Cisco rollout. This may well be one of the most difficult markets that Cisco's famed marketing machine has ever attempted to break into.
And it's not just a matter of control, though it is partly about control. It's also about how real-time applications behave in complicated deployments that cross the network owners' boundaries. As one of my No Jitter commenters noted, besides the issues of application performance and software licensing, if you build a service that's a hybrid of applications belonging to you, your service provider and multiple third parties, that is going to be a gigantic knot to try and unwind if you ever decide you need to make a fundamental change. We're talking about the ultimate vendor lock-in.
Having said all this, I can recall that when Cisco came out with its first IP-PBXs, people talked about how hard it was to do real-time traffic on IP, how primitive the technology was, etc. etc. And they were right. Until they weren't anymore.
So my bottom line is, this Cisco announcement is a big deal, and everyone needs to pay attention to it. But the technical and operational challenges are, to my way of thinking, many times more complex than putting voice over IP (which, though largely solved, still has its problems). The race between Cisco WebEx Connect and Microsoft Office Communications Server is a marathon, not a sprint.