We are a ShoreTel end-user for nearly two years, now, completely self-installed on the ShoreTel side of things. We have relied upon the valuable network expertise of our ShoreTel distributor to help us develop and learn the network strategies and skills required to keep VoIP happy. We don't call them much, any more, unless we find something that truly stumps us. Happily, that's an unusual occasion -- but it took a while. We have 29 sites in 4 states; 500+ users; 100+ analog devices (faxes, a few modems, etc). 99% of users have the IP530 or IP560 telephone. A wide mix of analog loop-start, analog DID, T1, PRI trunking. 3 or 4 of these sites exist merely to provide tie lines to existing Nortel PBXs (Option systems and Norstars), some using T1 and some using PRI tie lines.
Successful VoIP on the WAN runs over point-point T1s and MPLS. Early attempts at Frame Relay and Sprint's Sprintlink-Frame product just didn't pan out.
QoS is very important, of course. But it is amazing how critical a happy network is to VoIP communications. When call quality suffers, it's time to be checking error counts on every router involved, and even such simple things as half/full duplex negotiation on the network switches will have a huge impact -- and have to be looked at pretty regularly. And don't forget your bandwidth, you need your bandwidth.
The wise VoIP shopper will critically analyze his/her system and strategic needs -- and will dissect their network before implementing VoIP (!!!!). The wise VoIP seller will do the same -- it's your name that's at stake. In my experience, VoIP has some real benefits to the multi-site company. But it's more expensive to keep running: Your troubleshooters not only have to phone technicians (for the trunking issues, mostly), they have to be data technicians now, too (for the WAN and LAN issues), and PC technicians now, too (for the bells and whistles on the user PC), and know servers now, too ('cause it all hinges on the servers). Windows Server 2003 is more stable than 2000, but none of it runs forever.
You'd better have access to someone who can find that problem, and who can successfully argue and urge and politely insist that the carrier whose trouble it is get the darned thing fixed. The vast majority of our problems have been WAN problems. If your troubeshooter is a vendor, then I can imagine that money will need to change hands every time the WAN gets into trouble. The larger customer will likely hire his/her own in-house expertise.
A couple of years into this thing, I still believe the ShoreTel product really is a good product, with great new strategies available in the manipulation and delivery of voice telephony. The bottom line remains: If you are going to spend money on this business-critical technology (whether traditional telephony or VoIP), you had better do your homework. And don't be shy or hesitant in setting up a good ol' informal face-to-face with the folks trying to sell you this stuff. Get past the sales fluff, and figure out if you can trust 'em well enough to keep you in business. Same thing for you distributors -- we customers don't always know what questions to ask, and can be real intimidated by the monies and complexities involved, so you have to know how not to let the deal get set up for failure.
Hope this helps. Keep the discussion going -- it really helps to know what's going on out there.
