I've deployed quite a few Cisco based VoIP systems for small to large sized companies. I've also used Asterisk. There is no arguing with the fact that Asterisk is free. The debate will go on for years about the actual cost, or TCO of commercial -vs- open source, so I will spare everyone that.

What it really comes down to is risk. There is an old saying in IT, that "No ever got fired for buying IBM". In general (read 99.99% of the time), that is a true statement. Today you can say the same thing about Cisco.

When you suggest to management that you can produce a system with open source software that will have an over-all cost reduction of (insert your best guess at savings here), you are setting youself up to be either a hero, or a patsy, depending on how well everything goes. If you save the company $200,000.00 on their new phone system by sucessfully installing Asterisk, you will likely get a big pat on the back and perhaps, if you are VERY lucky even an extra grand or two in your bonus. If you suggest installing Asterisk to save money, and after several months the company's phone system is still dropping customer calls, or failing to route calls correctly, well you better have your resume up-to-date, cause the CEO is going to want someone's head. If your Cisco VoIP deployment goes all to hell, and you've kept Cisco involved when there are problems, then your butt is covered at least.

Astrisk is a great product (I use it at home), but I'm not going to recommend it for 5000 seat deployment, where if there are problems I can't find find fixes for, I have to blow the dust off of my C compiler to (try in vain probably) to fix. If I run into a problem with a Cisco deployment, I get support from someone that will write code fixes if necessary to fix my problem and get the system to work the way it was promissed.

BTW, can someone explain to me why "SIP rocks". There has been so much market focus on this over the past 2 years, and it you take the time to look at the fundamental differances of competing protocols, it just doesn't make sense (the hype I mean). The fact is that IP phones are pretty darn useless by themselves. They need to register to somekind of server or proxy, unless you want to enter your entire dial-plan into each phone. As such, they are really clients to a server. If phones are really clients, then why is the market so intent on using a truely (great) peer-to-peer protocol like SIP for phones. Don't get me wrong, SIP is great protocol, and much more fun to tweak on that client-server protocols like SCCP, but using SIP as a determining factor in choosing a VoIP solution is like using the color of the face bezel to determine what servers you are going to buy. The fact is that Cisco supports SIP, though in 95% of the deployments I've worked with it's never used or needed. And SCCP is supported by Asterisk, but most Asterisk deployments use SIP without ever missing the easy setup and management that SCCP offers, because Asterisk uses other ways to handle those tasks.