For me, the recent VoiceCon show in San Francisco gave new meaning to the words "unified messaging."

As I made my rounds to close to two dozen analyst meetings, almost every executive was focused on laying out their company's "Unified Communications" strategy and/or its upper-stack cousin, "Communications-enabled Business Processes." UC and CEBP were certainly the stars of the show but how we'll get there is not at all clear and a big battle is brewing.

IP telephony battles of recent years have, naturally, been between the long-standing PBX vendors - Alcatel, Avaya, Nortel, Siemens, and the "new" VoIP vendors - Cisco, Shoretel and a string of others. Now, with its Office Communications Server 2007, Microsoft is arriving in a big way and has big plans to take over IP telephony, er, sorry, unified communications.

In an hour-long product commercial, apparently mislabelled as a keynote speech in the event programme, a Microsoft executive spoke in detail of how the aforementioned OCS 2007 and the client-counterpart Office Communicator 2007 would, essentially, eliminate the need for old, "hardware-based" systems. Meaning, in essence, anything sold by anyone other than Microsoft.

As I sat there, I couldn't help but remind myself that there is really no such thing as a "hardware-based" PBX any more. Years ago, PBX systems did run proprietary software usually on proprietary hardware but those days are over. The "traditional" vendors have all ported the most important system elements to run on open hardware and OS platforms. All offer "soft phones" that run on popular PC clients and VoIP "hardware" phones are more software than hardware.

As it happened, my next meeting after this session was with executives from Avaya. When I asked them if Microsoft's PBX-elimination strategy concerned them, they said it did not. Why? Because they believe that CEBP will trump unified communications. To explain: Where unified communications provides integration of "general" functions - like being able to call someone by clicking on their name in Outlook - CEBP will provide a more significant value-add.

According to Avaya and others, CEBP will allow communications tools to be directly integrated into business processes. Avaya's example (and another keynote topic) was Black and Decker. CEBP (from Avaya) allows the company to use text-to-speech to allow computers instead of people call out to customers whose products have been repaired and are ready for pickup.

While it is hard to argue with any of the individual approaches, they can't all win. In the past, the battlefield is usually limited. For example, when IP telephony came about the struggle was between old-line PBX "telephone" departments and the IT department. With server virtualisation, the struggle is often between the server people and the data centre infrastructure teams.

When it comes to unified communications/CEBP, the battlefield can run all the way from the IP telephony team, across to those responsible for the company messaging and server strategy all the way up to the application teams that program the company's line-of-business applications. And, it might be that where the battle takes place will determine the winner.

If the "CEBP" strategy can truly have the application drive the communications technology, then the traditional telephony vendors have a good shot. All too often, though, the IT infrastructure is put in place for subsequent use by application teams. If that happens this time, it could be an all-Microsoft communications world before anyone notices.