Q&A with In-Stat on a new study: IP PBX adoption rate dependent upon new services, software
Director Joe Pavlat spoke with Norm Bogen, Director, Networking Research at In-Stat about the IP PBX market, the applications that will drive companies toward “all IP” based systems, and the role of CompactPCI, AdvancedTCA, and MicroTCA as this drive takes off.
Q: What is the 50,000-foot view of the global PBX market?
A: For 2006, we are looking at 9.5 percent growth in line shipments, with the total PBX market growing to about 25.5 million lines by 2006. Of that, 7.4 million are the traditional TDM PBX, and 18.1 million are IP PBX.
Growth overall is 9.5 percent, with growth for IP PBX at 14.3 percent, and there is going to be negative 0.8 percent growth in the TDM PBX, the traditional PBX. The market will experience fairly rapid growth in IP PBX, to 37.5 million IP PBX lines in 2010, which is 98.8 percent of the total.
We’re moving to all IP, in terms of what is shipping. Let’s stop a minute and talk about the installed base. When we look at the installed base today, many companies that employ IP PBXs, especially for the enterprise, have deployed an IP PBX somewhere in their company, and they may even have deployed several. But most of these large enterprises continue to maintain their TDM PBXs, so they are using both for now.
There are a few companies (most of them are CISCO customers) that have trashed their TDM PBX company-wide and replaced it with all IP-based systems. Cisco has made these replacements, but not very many other firms have. Most people are continuing to deploy their existing PBX to get their return on investment until its end of life.
And in fact, our company has asked about ‘going IP,’ but no huge cost savings on long distance calls takes place with IP to motivate that change.
Our advice to our own company is: “Continue to use your existing PBXs. Experiment with an IP PBX in a location where you are going to replace it. When it is time to replace a PBX anyplace in the company, throughout the world, don’t replace it with a TDM PBX, replace it with an IP PBX.”
The other piece of this is the IP phone itself. With a lot of these companies, when they purchase an IP PBX (and again this relates to what we recommended to our company) there is not a driving need to replace the actual handset, because that is the most expensive part of the system. So most companies are not replacing it yet.
What we found in 2005 and 2006 is that for those who deployed an IP PBX – about 40 percent of those lines went along with an IP phone and about 60 percent of the IP lines were shipped to companies keeping their existing handset.
And again, over time, more IP phones are going to displace the TDM phones. As those TDM phones need to be replaced, companies are going to replace them with IP phones.
Q. I have seen new office buildings in China with only one pair of wire going to each desk that will be used for voice and network access.
A: I think what we recommended to our company is pretty much what most companies are implementing. We are forecasting that in 2010 98.8 percent of all PBXs that ship will be IP based. Yet even in 2010 a significant number of companies will continue to use TDM – I do not know the exact timeframe, but it is going to take 10 years, maybe 15 years, to replace all of the old TDM systems.
Q: Do you see this happening across the board for companies of all size ranges?
A: It is occurring across the board, but it is mostly now deployed by companies with 1,000 or more employees. The smaller companies have not typically yet done it.
Q: How much is open standards- based from a hardware perspective and from a software perspective?
A: Most of it is on open standard hardware.
Q: Except Cisco.
A: Right.
And software-wise, most of it is not. The players – Nortel, Avaya, Cisco – want you to buy their handsets. So they are selling you their software that works with their handsets.
Q: What percentage is going on standard rack-mount servers versus CompactPCI?
A: I do not know the percentage, but I believe that most of the stuff everybody but Cisco is shipping is a commodity server.
There is also a new product category, on which we are doing some research, called a business gateway.
Cisco has a business gateway product and again it is proprietary, but others – Samsung Telecommunications, Avaya, Alcatel, Nortel – all these firms are developing products based on open hardware standards that we call a business gateway. It includes:
Router
Switch
IP PBX
Firewall VPN
Potentially supports wireless LAN switching as well
The business gateway is all integrated into a single box. The question is: could this single box be CompactPCI? I know it could be AdvancedTCA, but that is overkill. It could be MicroTCA.
I have been talking to Intel about this, saying, I think there is an opportunity for an enterprise class product using AdvancedTCA standards.
This class of product does not need NEBS III compliance and 5-nines reliability, so AdvancedTCA may not be the answer.
Q: So this business gateway could be an opportunity that incorporates the PBX, but getting back to the PBX market, is that a place for CompactPCI?
A: When we compare CompactPCI, say, to a regular HP server, we’ll find they generally are very similar in function, but CompactPCI has a little better density, a little better cooling, and the opportunity to have redundancy and hot swapping if needed.
Also, there are versions of CompactPCI that have Ethernet switches in them, and all the backplane traffic is Ethernet based. It’s a more compact way of putting bunches of servers and switches in a big rack.
Certainly CompactPCI has a role in the TDM world, for example, with companies like NMS Communications – many of their voice products are based on CompactPCI. Their older products are TDM, but they are making the switch to IP also, and CompactPCI supports an IP backplane.
Although the NEBS compliance [CompactPCI offers] is not needed, reliability is a plus because people do not like it when their phone system goes down.
What’s more, for applications hosted in a telco environment, the more reliable hardware platform makes sense.
There is an installed base of existing customers of what is called Centrex, which is a hosted TDM solution. There is an IP-hosted solution as well. And in the IP solution, I think CompactPCI again, makes sense, and may even be the bigger platform, because there are a couple of ways to go with this:
1. Telco hosts everything for your company on their server in their building
2. Multiple clients on one box, and that might be a bigger box
In companies with less than 1,000 employees is where we are seeing green field opportunities. In addition, there is an interesting battle between these lower cost PBXs, maybe even open systems, that a company would buy, which is in our report The Changing Face of IP PBXs: Mobility and Multimedia (Report # IN0602853CT), versus the hosted solution. With a hosted solution, the telco is going to provide that service because the smaller company does not have the expertise in the IT department to handle it.
Q: Is there an opportunity for smaller companies that are often virtual and geographically spread out? Could employees in many disparate locations benefit from a hosted service?
A: Absolutely, and there is tremendous savings and ease of use as well.
Q: When IP hosted based phone services cost me less than 1.5 to 2 cents a minute as you suggest, will IP-based telephony dominate?
A: Yes, and then you’d have three- or four-digit dialing to all your compatriots.
Q: For many of the hardware vendors, the question is not CompactPCI/TDM versus IP, it’s CompactPCI versus proprietary hardware solutions and standard servers.
My guess is that the majority of standard servers are not CompactPCI. There are, however, companies like NMS that do use CompactPCI because they are selling a particular solution to a particular market.
A: I see the competition between CompactPCI and standard servers for the IP PBX market.
Q: Sure, and the server market is largely an enterprise market. The traditional mobile suppliers – and everyone is talking about triple and quadruple play – are competing heavily with some of the traditional telco service providers. Are they going to offer converged services?
A: They already are. Where it really makes sense is for Fixed/Mobile Convergence. The wireless operators have to cooperate with somebody when it comes to this – Wi-Fi/cellular combo handset that we have a big forecast for.
Doing seamless handoff from your enterprise network to the mobile network requires that cooperation, and that’s where we think there is going to be a lot of demand.
Q: Everyone wants one phone number and they want it to follow them around wherever they are. We’ve been hearing about one number follow me for years, but it doesn’t seem to be reality for most of us.
A: The carriers operate on pretty slim margins. There is strong competition for added-value services.
If commodity servers will do the job, most customers will tend to go that route because they are built in huge volumes and they are dirt-cheap. You’ll get things like CompactPCI, which is a little higher performance, but a little more expensive technology used in more niche-y applications.
Telephony is one application where you may want higher reliability.
Q: What about military or government customers who appear to want a higher level of reliability and security?
A: Security is a big issue, especially when you go wireless.
Q: Security is an issue for everybody. Security is a huge issue in wireless and in VoIP. As more people roam between networks, and as more people use VoIP, it is going to become a much bigger issue.
In some cases, the military is already using CompactPCI because standard servers are not tough enough. Net-centric warfare and the concept of any data, anytime, to any soldier, anywhere, as well as the whole infrastructure behind that, is not a whole lot different than a teenage girl yakking on a cell phone, and so I would expect that there will be niches where CompactPCI would be a very neat solution for shock, vibration, temperature, and the like.
A: MicroTCA might end up being a good solution for small- to medium-sized enterprises where you’ve got something [MicroTCA] that gives higher reliability than a standard server. It may not be CompactPCI, but I think as MicroTCA matures, that MicroTCA is going to be a logical candidate for enterprise-class IP PBXs. AdvancedTCA is probably overkill.
You want an IP PBX throughout your company if you want to take advantage of new functionality, for example, presence. Also, collaboration, conferencing, and other IP type applications, and even video are easier to do.
All these new applications are going to create demand for companies to dump the TDM system and replace it with all IP. Watch the applications that are coming out that are going to create demand for all-IP systems.
Norm Bogen is director, networking research at In-Stat.
