Slashdot Mirror
Cheap and Reliable IP Telephony?
anomalie asks: "I am trying to sell IP telephony to my employer. The idea was shot down once already because of the cost (using a Cisco solution). I would like to find a cheap but reliable IP PBX because everyone liked the idea of IP telephony, just not the price associated with it. I need a system that could initially handle about 80 users at a single location, and eventually handle about 350 users at 7 locations. The two systems I have been looked at so far are Asterisk & Pingtel's SIPxchange IP PBX. I'm not looking here for a final solution, just some starting points for more research. Any feedback/tips/warnings from the Slashdot community?"
"I am looking to have at least the following capabilities:
-Auto attendant
-Handle a PRI (hopefully allow forwarding of old PBX DIDs)
-Handle long distance T1 (we would initially segment off some channels from our current PBX)
-Handle WAN Traffic so we could utilize our unused channels for long distance from other locations
-Forwarding of voicemails to email
Nice optional features:
-Web based GUI for voicemail administration
-GUI call manager
Eventually, we would have relay units at the other locations to handle the local calls and call routing and have 1 central PBX at corporate headquarters."
-Auto attendant
-Handle a PRI (hopefully allow forwarding of old PBX DIDs)
-Handle long distance T1 (we would initially segment off some channels from our current PBX)
-Handle WAN Traffic so we could utilize our unused channels for long distance from other locations
-Forwarding of voicemails to email
Nice optional features:
-Web based GUI for voicemail administration
-GUI call manager
Eventually, we would have relay units at the other locations to handle the local calls and call routing and have 1 central PBX at corporate headquarters."
It seems like replacing your current phone system is going to cost more in the short term than just sticking with what works. Will the IP telephone system cost less in the long run? What is the time frame to break-even? Will your company still be around by that time?
I have been pwned because my
The only corporations that would actually stand for a gain from the benefits of IP phones are the ones that would not balk at the price of the hardware.
The largest sum of money spent on phone systems is usually interoffice calls. Why would you set up IP phones in a single location as your goal, then add more offices as a secondary "good to have?"
This is not sound economics -- just because speaking over your cat5 network is cool does not make it a smart thing to do.
...
I work in a company with IP phone system. While it's good for long distance calls to branches in other cities (pay local call rates), the overall system has let us down many times.
I'm not sure whether it's that this particular service provider is no good, or whether the service itself is still unproven.
Anyway, just remember when the phone doesn't work, your internet also doesn't work.
No phone calls & no emails - might as well go to the bar.
Phones and emails are almost at the core of most businesses now, they are expected to be always working (like electricity in the building), and when they don't work, the managers get really upset.
Anyway, I don't know who (this person is probably no longer with us) got the company to use IP phones, but they have mentioned many times how much they hate the system.
Good luck. May be services in your area are much better.
When you looked at the Cisco solution, did you just look at the bottom dollar or did you run the numbers on line savings, labor savings, etc? We're going from an all-Centrex phone system with about 420 lines spread across 21 sites to a Cisco VoIP system in the next few months. The initial cost is pretty high, but we're looking at a 2-2.5 year payoff. After the payoff, we're estimating a $7-$8,000 savings per month! Our network techs can do all the maintenance (no more $200+ service calls), all phones have advanced calling features and voicemail, we're leveraging the XML-based displays to add company directories and clock-in/out capability, etc.
Don't just look at the actual cost, run the numbers on everything else too. If it doesn't work out now, stick with current setup until either the numbers work out better or until the added features justify the cost.
Also, I personally wouldn't want to stake my company's phone system on a smaller vendor. We looked primarily at Cisco, 3Com, Nortel, and Avaya. All three have good reputations in the industry.
Jason
"FORMAT C:" - Kills bugs dead!
Moving to IP based phones can have some great benefits to your company. If you do it right you will end up:
- With almost unlimited flexibility for managing call routing
- Easy integration with databases for tricks like skills based call routing
- Low cost for intra-office communications
- Near perfect support for work from home employees
- Flexibility to set up spot call centers and so on
- Really cool voice mail
Unfortunately, most companies don't use a fraction of the capability of IP phones - a lot of time they end up being used just like the crappy 80's AT&T Merlin they are replacing. Oh yeah - make sure you have a really good SLA on your T's... downtime is SUPER EXPENSIVE when sales and customer service are down.
-- $G
Wow there are so many armchair quarterbacks in here it is unbelievable!
I work at a mid-sized telco that heavily relies on IP telephony. To put it simply, this is where things are moving to on the carrier side and the PBX side. The technology is mature. Everyone is using it. 'Nuff Said.
It looks like the poster is looking for a basic IP PBX that does the stuff that pretty much EVERY modern office PBX does. AutoAttendents, lite web client, simple IVR's, voicemails, and being able to interconnect with a T1 are all very standard features.
Having researched the PBX and call center solution for my own company (about 300 users with 100 call center agents), these are my 2 recommendations:
Artisoft Televantage
--------------------
-VERY Inexpensive for small offices like 5-20 people. Pretty average priced when you get up there in the users. Low base cost, high per-seat license cost.
-Supports pretty much EVERY feature under the sun, along with some neat stuff like 'follow me' routing lists, announced hold times, and a free SDK for ODBC integration if you want to build your own IVR's and plugins.
-Televantage runs on standard Intel Dialogic boards, so you can use T1's, DS3's, POTS lines, whatever you want. It also supports something like 1000 SIP users per server if you want to use standard SIP IP phones.
-Biggest disadvantage to TeleVantage is that it runs on a lite version of MS SQL server. On average, we reboot our TeleVantage system about once a month just for stability's sake.
3com NBX SuperStack
--------------------
-The 3com is pretty lite on features though it does cover everything that the poster asked for. Certainly not a solution for a call center, but defenitely a great box for an office environment.
-The 3com box runs Cisco Call Manager which is a plus since the poster specifically said he likes the Cisco stuff.
-The 3com box is very inexpensive for small-to-mid sized offices of like 30-50 people. The license cost and the base cost are both reasonable.
-The 3com box runs the same OS as artificial hearts, so it is VERY VERY VERY stable.
-Disadvantage is that you have to use proprietary 3com phones since insted of going with a standard protocol, 3com uses some Layer2 ultra-efficient monster of a codec that they developed internally.
-Another disadvantage is that if you want to add features that are not available in the 3com SuperStack, you basically have to put them on a seperate box next to the machine. For instance, if you desperately wanted ACD or announced hold times, you'd end up putting a 2nd box just as expensive as a PBX right next to your PBX to handle those calls on pass-through.
I second that.
I am using Asterisk right now to offload our high-volume long distance calls over to Nufone. $0.0295/min anywhere in the continental US and Canada. Great service but not offically open for business yet. Talk to Jerjer in #asterisk on OPN.
Anyway -- Asterisk for the most part works great -- I currently have our Norstar system with four trunk lines going into an Adit600 channel bank to Asterisk. We also have 8 regular PSTN lines which go directly into the KSU. Speed dials are set to pick up a VOIP trunk line. When we move to the new building we will have a PRI going directly into the asterisk box, and a channelized T1 connection between the Norstar system and the Asterisk box. We're only going to have one "real" phone line in the building, with everything else going over VOIP to a colocation place downtown.
Biggest problems with Asterisk (for us) seem to be with VOIP phones, not VOIP calls. Since we're using our regular Norstar system we avoid most of these issues but we are slowly moving to VOIP phones to replace the KSU since we want (much) tighter integration of the phones and computers. You pretty much require end-to-end QoS though for guaranteed reliability and clarity of calls. We do pretty go with having QoS working on both ends of the data T1 such that it's not possible to fill the pipe and cause havoc.
Asterisk is really becoming VERY stable over the past few weeks -- I think there are under 8 bugs open in the bugtracker which are preventing 1.0 stable. (yes I realize how funny that sounds)
The company I was in previously used the expensive, Cisco system (and it was cool to see our phones in the movies - Ocean's Eleven and others)... and we still had issues with it. Even once all the wrinkles were worked out and we no longer had echoes or constant rebooting of the phone server, it was still only as resilient as our network.
Like parent poster said, it sucks when your network goes down and you reach for the phone only to see that it's also rebooting, and you are stranded.
Granted this was a year and a half ago, but you're still taking the risk of much greater technical complexity, plus sharing a network that can be brought down by a lot of other factors (whereas POTS is independant).
Before you get on the boat, you'd better be able to point to a significant savings to justify it AND you have to either factor in downtime, or pay for a bank of backup standard phone lines. Here's a good tip for evaluating providers -- ask them for contact info for a few current customers that you can talk to. They should be able to find one who can share the experiences so far. You do NOT want to be the guinea pig on the cutting edge.
There are only 10 types of people: those who understand decimal, those who don't, and, uh, 8 other types I forget.
The company I work for recently (~6 months ago) sold a system almost exactly as you describe, to great effect. My job was to make it all work together between them closing one PM, and reopening the next AM. So far, they're quite happy with it.
As configured, it has an LD T1, local DID PRI, auto attendant, VM retrieval by email, slick client-side GUI, about 120 analog (POTS) extensions, a handful of active h.323 IP extensions, and an operator console. Consumes only 4 rack spaces, instead of the couple dozen square feet of wall space occupied by their old switch. 17 PCI slots, hotswap Adaptec RAID, hotswap redundant power supplies, redundant quick-connect fans, audible alarms, gig-o-RAM, backplane, captive screws, yadda, yadda.
80 extensions in one spot, be they IP, analog, or 80 of each, is not a big deal.
Runs Win2k, has a network-operable Win32 GUI for administratia. It'll do all the fancy automatic least-cost call routing you can ask for between branches (via IP, or whatever other means you have). It will also do the remote PBX thing at least as well as anything else available today. Tenant-oriented resource allocation and detailed call reporting (and recording, if that's your gig) will keep the beancounters happy.
It's called Altigen. It mops the floor with Cisco's paltry offerings, across the board. And it's way, way cheaper.
Questions?
Kid-proof tablet..