Optical/IP Networks

Enterprise VOIP: Not So Fast

I've got to admit, I'm starting to see the attraction of voice over IP (VOIP). As a Skype user, my experiences have done quite a bit to reform my inner VOIP cynic. And it seems I'm not alone: Market research companies are also forecasting a huge boom in residential VOIP (see Residential VOIP Will Boom, Says Study).

But convincing the wider business market is a different matter altogether, and it raises a couple of key questions:

  1. What's happening to enterprise VOIP? Is the residential market blazing a trail in VOIP that business users will follow, or is the opposite the case?

  2. When and if the enterprise VOIP market takes off, who will be the big beneficiary?

Here are my thoughts on Question No. 1: There have been plenty of announcements of enterprises taking the plunge on VOIP. My advice is to take those reports with a pinch of salt.

In most cases, these announcements are really toe-in-the-water type trials that vendors are attempting to pump up, in the hopes of convincing other enterprises to make a move in the direction of VOIP.

The truth is that rolling out VOIP networks in enterprises is a totally different cuppa tea from an individual (like me) choosing to download and use Skype.

On the one hand, VOIP holds out the promise of delivering a lot more than a replacement of PBX networks – such as presence services, unified messaging, and conferencing. For more on this, check out SIP Hosted Services: A Heavy Reading Competitive Analysis, the latest report from Light Reading's paid research division.

On the other hand, enterprises aren't going to scrap their huge investments in PBX networks overnight. They also face all sorts of other challenges – notably, having to upgrade their internal IP infrastructure to carry VOIP and dealing with firewalls and network address translation (NAT) schemes that make it difficult to receive incoming VOIP calls.

The business case for rolling out VOIP in enterprises also isn't clear-cut. It's true that IP PBX phone prices have come down. You can now get an IP PBX system for between $500 and $900 per user, including phones. But the true cost savings are only there for enterprises willing to go all out with VOIP across their entire company, and that's a big commitment to make, as already noted (see Heavy Reading Sees Money in SIP). LAN upgrades can also hack at any savings, bumping up the cost of a new IP PBX system significantly.

So, Question No. 2: Which companies are well positioned to cash in on enterprise VOIP, if and when it really takes off?

I've just completed a study entitled "VOIP: The Enterprise Options" for Light Reading Insider, the subscription research service of Light Reading, and here's what I think:

The lion's share of IP PBX customers will opt for CPE-based equipment and will likely choose between products from Cisco Systems Inc. (Nasdaq: CSCO), Avaya Inc. (NYSE: AV), and Nortel Networks Ltd. (NYSE/Toronto: NT).

LAN dominance counts for a lot, and Cisco, the current IP PBX market leader, holds the market by its short and curlies, only guaranteeing service-level agreements (SLA) on its own LANs. Other companies are forging their own LAN/IP PBX relationships, with Avaya and Extreme Networks Inc. (Nasdaq: EXTR), for example, forming a partnership intended to rival Cisco's single-sourced advantage (see Avaya & Extreme Team on VOIP).

Avaya and Nortel, the legacy PBX market leaders, will also naturally prosper from their installed base and are cementing their presence in the IP PBX market with new products to match the requirements of enterprises of different sizes.

Should an enterprise take the uncharacteristic step of outsourcing, there are a number of hosted IP PBX and IP Centrex providers, using equipment from vendors such as Sylantro Systems Corp. and BroadSoft Inc., that are gunning for enterprise dollars.

But will enterprises really overhaul their legacy of in-house PBXs with an outsourced option? I think not. The cost-sensitive SMB market will most likely adopt hosted services in the short term. Enterprises, meanwhile, will deploy IP PBXs in tentative, measured steps that will only outweigh TDM beginning in 2007. It's anyone's guess which vendors will benefit as enterprise VOIP becomes more commonplace, but I wouldn't be surprised if Cisco remains king of this castle.

— Daro Clark, Research Analyst, Light Reading

To subscribe to Light Reading Insider and obtain a copy of its latest report, "VOIP: The Enterprise Options," simply click here.

"SIP Hosted Services: A
Heavy Reading Competitive Analysis," a 56-page report by Margaret Hopkins, costs $3,495. For more details, click here.

Page 1 / 2   >   >>
gbennett 12/5/2012 | 1:47:23 AM
re: Enterprise VOIP: Not So Fast Comrades,
Those of you who've read my occasional rant about VoIP on these boards over the past few years know that I've been a skeptic for a long time. But, as Daro mentioned, we've been dabbling with Skype recently and I was amazed at how effective and simple it really is. A friend of mine called me on Skype this morning using a dial-up Internet connection from his hotel room in Germany! The call was clear as a bell - a bit more delay that I'd heard on other Skype calls, but no worse than a lot of cellular calls I've experienced.

I haven't tried FWD or the other options, so I'm sure they're good too. But the PC-based phone approach has clearly got some rough edges. For example, it's be great to have a DECT (in-home wireless) handset or headset so I can continue the call while moving around.

It would be good if I could plug in my headset (I use an old voice recognition headset I bought ages ago) to the laptop, but the ringing noise would still come out of the speakers.

It would be good if there were more than 5 people I could talk to :-) But that will change I think.

I don't see Skype replacing my phone just yet. But I'll be making as many calls this way as I can, and will save a ton of money (both personally and for the company) in the process.

I can see now why Jeff Pulver has been so evangelical about this - you really need to try it in order to appreciate just how scared the incumbent voice operators should be :-)

lastmile 12/5/2012 | 1:47:22 AM
re: Enterprise VOIP: Not So Fast When someone like 'lastmile' says that skype is great, they call me a propeller-head.

When gbennett from LR says the same everyone listens.

Keep it up LR.
gbennett 12/5/2012 | 1:47:00 AM
re: Enterprise VOIP: Not So Fast Hi Lastmile,
Was that "everybody listens" or "everybody laughs" :-)

There's no doubt in my mind that Skype, and I'm sure other "personal VoIP" packages, are revolutionary. But the darn thing is I still can't call my mum on Skype, or most of the friends we know. Having said that, word of mouth is the best advertising. I converted about a half dozen friends at a part we went to Saturday night. Most of them have relatives in foreign parts, or kids in college, etc.

So Skype is revolutionary within a subset of the calls we make. BT will still get my monthly line rental payment, but an ever-diminishing revenue from actual calls I would hope.

As I understand it, the voice quality of Skype is better than most, is that right? I'm stuck with ISDN until August (can't come soon enough), so I can't seem to get Instant Messenger "talk" features to work very well because of the bandwidth issue. But Skype works great over my poxy 64kbps B channel. Couple that with the ease of installation, and the fact that I didn't have to mess with my firewall, and it points to them getting a lot of things right.

dljvjbsl 12/5/2012 | 1:46:59 AM
re: Enterprise VOIP: Not So Fast It is a simple fact now that the replacement market for enterprise voice is the IP PBX. As enterprise TDM PBXs reach their end of life, they are being replaced by IP PBXs. The market share for IP PBXs is growing at double digit rates and fromthe numbers that I have heard will replace the existing TDM market by 2007.

Hosted PBXs and IP Centrex have one major problem. They are forms of Centrex which enterprise customers know and by knowing, hate.

They hate Centrx because of high prices for routine work, and minimal and out-of-date feature sets. The telcos have been proclaiming the rise of Centrex and the end of the PBX for decades. However in all that time Centrex has remained the uncompetitive service that it is.

The telcos are not going to change.
PO 12/5/2012 | 1:46:57 AM
re: Enterprise VOIP: Not So Fast They hate Centrx because of high prices for routine work, and minimal and out-of-date feature sets. The telcos have been proclaiming the rise of Centrex and the end of the PBX for decades. However in all that time Centrex has remained the uncompetitive service that it is.

Herein lies the rub which I think the article correctly addresses: where is the breaking point between internalizing and outsourcing tech infrastructure management? The line keeps moving as new technologies both simplify and complicate the task. And as the companies start small and grow to medium-sized and large companies.

How well the carriers compete in the IT outsourcing game will go a long way in determining their future shape, as VoIP becomes just another task for Information Systems groups.

Then again, this is all IMHO.
fgoldstein 12/5/2012 | 1:46:43 AM
re: Enterprise VOIP: Not So Fast Much of the positive press on parasitic VoIP comes from people who tried it and initially liked it. Hmmm, that reminds me of something. Ever hear a junkie talk about their experience? They tried it, they liked it, and the world didn't end right away. Then they were hooked; when the bottom did fall out, they were in real trouble.

Parasitic VoIP can work pretty well. But it's like an English car: Fun when it works, but don't count on it. Since there is no reserved bandwidth, the call gets through by the grace of the available bandwidth not being heavily utilized at that moment. So the test drive is great, but you may have second thoughts later. (In Skype's defense, though, the price is right.)

I visited a friend over the weekend who had tried Vonage. He had to drop it, because the other end couldn't hear him speak whenver he was uploading stuff from his computer. His complaint was about the cable modem network, naturally, for having too low an upstream cap. He didn't quite realize that packet discards are normal in IP networks. He had the newer Motorola adapter, which does voice priority, but that didn't do the trick.

Skype and Vonage and their ilk can give you a nice high, but they are not a good long-term way of life. At least not if you depend on them for reliable telephony. That kind of VoIP is, however, a good CB simulator or voice-chat system. It's useful for casual conversation, but probably not lifeline or business. PacketCable is a different animal altogether; it has reserved bandwidth (from DOCSIS 1.1+), so the voice and data are, effectively, separated below the IP layer. But of course cable ops aren't giving that away for free.
gbennett 12/5/2012 | 1:46:31 AM
re: Enterprise VOIP: Not So Fast Hi fgoldstein,

Not sure I like the English car analogy (probably 'cos it's true :-) , but I do agree with you.

I don't see Skype as a phone replacement. But actually it's fine for business calls - in some ways better and more conventient than a PSTN link because it's driven from the PC, so messaging is included, there's a contact list etc.

My future phone bills will keep the line rental component, but the amount for chargeable calls will drop dramatically I suspect. And that's bad news for my telephone provider.

[email protected] 12/5/2012 | 1:46:31 AM
re: Enterprise VOIP: Not So Fast IGÇÖm not too sure about the drug analogy. Perhaps the residential VOIP we have now is a GÇ£gatewayGÇ¥ drug that introduces the opportunities that GÇ£harderGÇ¥ options offer, but itGÇÖs not quite hit dependency levels yet. (Then again, I could just be an addict in denial)

The key thing for me is that VOIP services like Skype will go someway to convincing the naysayers who have previously not given VOIP the time of day (I know of plenty already) These attitudes will filter through to business users, who will be more willing to pay for a VOIP service, especially as functions and features increase.
aswath 12/5/2012 | 1:46:29 AM
re: Enterprise VOIP: Not So Fast who will be more willing to pay for a VOIP service, especially as functions and features increase.

My point of departure with some ofthe VoIP proponents is regarding the aspect that I have highlighted in this quote. As I see it true P2P is the only way to deploy VoIP. The service provider model will have distinct disadvantage over PSTN for a long time to come.

All the applications and services (including those mentioned in Msg. 7) can be realized in PSTN as well, if we grant computing power and display devices to the end-points. For example, Messaging and contact list is included in the telephone application that came with my Gateway computer.
fgoldstein 12/5/2012 | 1:46:27 AM
re: Enterprise VOIP: Not So Fast I'm sure that P-to-P VoIP will cost the telephone industry something, but "free" bandwidth is still more of an accounting trick than reality.

What's clear is that the telephone industry's 75-year-old "station to station" accounting technique is obsolete. (This dates back to the 1930 Smith v. Illinois Bell decision of the US Supreme Court, which held that local loops, being used for interstate calls, were partially under federal jurisdiction. Other countries found other means of using LD to subsidize local.) Designed to subsidize (non-luxury) local telephone service from (luxury) long distance revenues, it was useful for decades, helping promote "universal service", but got out of hand years ago. What parasitic (i.e., using bandwidth paid for for a different purpose, namely data) VoIP services provide is, in effect, "board to board" accounting for bandwidth. No contribution to the local loop.
Which quite frankly is fine, if one accepts that Smith is obsolete or at least needs to be handled differently.

But the bandwidth is not free. It's a step function. You buy x bps of Internet service, and if you don't use it up, it's fine, but if it gets full, you need to buy a bigger pipe, at a much higher monthly price. Telephone service, on the other hand, is paid for by the call. More visibility, to be sure. But not all that different in practice. Remember WATS lines? Calls were "free", but telecom managers had to size up their WATS groups. The net per-minute rate was only a modest discount. (Been there, done that.)

The less that the subsidy system distorts things, the more accurately people will choose technologies on their merits. Right now VoIP benefits from arbitrage. It will still have a market without arbitrage, but remeber why Vonage shouldn't be pronounced VON-idj, but voh-NAZH. Even if they don't go along. ;-)

BTW, y'all, the FCC's VoIP Docket closes in a couple of weeks, and not many comments have been put on ECFS yet. (Look for "04-36".)
Page 1 / 2   >   >>
Sign In