The business case for VoIP
Benefits of VoIP
Where exactly do the benefits of VoIP for businesses lie? Much depends on your current situation, current voice requirements and future plans. These scenarios aim to highlight the various considerations:
1) Moving to a brand new office
Go with VoIP. It's that simple. If you don't have any pre-existing infrastructure to worry about, fitting up a new office is the perfect opportunity to implement VoIP, regardless of your voice requirements. Why? You'd have to fit out a network infrastructure anyway, no matter how basic. If you followed the traditional model of voice/network infrastructure, you'd have to cable everything up for network communications AND lay a separate run of cabling for voice communications. Not to mention the cost of either a dedicated line to the exchange, or a trunk line and local PBX. Consolidate and save.
2) Existing single office, low density, no PBX
This scenario fits the description of only the smallest of offices with perhaps only one incoming line, much like a residential connection. The benefits of VoIP here are not clear-cut. If you have an Internet connection, it would certainly be worth considering approaching the ISP to see what broadband VoIP deals they can offer. This would give you the flexibility to have multiple handsets on the one line or even a VoIP-capable mobile running off a wireless network. However the decision is more likely to be made on the grounds of functionality and consolidation rather than cost. The cost differential between traditional voice and VoIP is likely to be negligible.
3) Existing single office, local PBX, trunk to exchange
Are your current voice requirements being met? If so, stick with what you have for the time being until it's time to upgrade, and then future-proof your requirements by spending the money on VoIP. Until that time it's unlikely to be worth the cost. If you're not satisfied with the current system and were looking at upgrading anyway, VoIP is definitely a good move. You'll still need the trunk to the exchange for PSTN connection, but you'll get the internal benefits of running VoIP.
4) More than one office, local PBX at each, multiple trunks
This is a scenario where VoIP truly shines. Each office has a requirement for a local PBX to minimise costs and meet internal voice requirements and as already mentioned, even a VoIP-enabled system eventually requires access to the PSTN. But what you can do is leverage off each office's Internet connection to set up VPNs across the WAN links between each office, effectively making them part of the same network. Upgrade each office to VoIP, have one office hosting the trunk to the PSTN, and route all external calls through this one connection. Slash costs and make every employee's extension effectively internal (free!). Even as a starter, you could keep the analogue handsets but get rid of the trunks. Very, very cool.
Problems with VoIP
Before you get too excited about VoIP, it's worth taking a moment to consider what can go wrong. Although the beauty of VoIP is that it moves you from a rigid, centralised, hardware-based voice system to a decentralised and flexible one, this same selling point is also the source of many of VoIP's problems.
As soon as you move to a VoIP system, you're now running all your business communications over the network. This means that the bandwidth used to access databases, work files, e-mail and the Internet is now sharing the same space as your voice comms.
At this point, quality of service becomes paramount. The older phone system never had much of a problem with QoS, as each phone line was essentially a dedicated circuit -- next to no chance of interference. Now, the VoIP network packets are mixed in with everything else.
Network lags are common enough problems. Most of them don't cause much drama -- a file takes longer to copy, a Web page takes longer to load, a server share takes longer to browse. Most users simply take this in their stride, and it takes a severe lag or series of lags for complaints to be raised. However from personal experience, I can guarantee that users are far less understanding about phone problems. I think it's because computer communications like e-mail and IM are still regarded as inferior because they are impersonal -- no emotion or voice inflection, risk of misinterpretation and so on. Many people still rely on phone conversations as their primary means of communication, but even those who don't won't tolerate lags, echoing and drop-outs while they're on the phone. The PSTN has set a high standard of quality over the years, and VoIP has to at least match it.
So with network reliance paramount, you have to ensure that your infrastructure is up to the job of moving all that data around. In scenario number four, where VoIP replaces multiple trunk connections, routing and switching across the WAN has to streamlined, which is even harder than optimising the internal network. And optimisation is the absolute key to VoIP. The vast majority of technical problems with VoIP relate to the supporting network infrastructure. Get it right first time and you can run anything -- get it wrong and it will become the bane of your existence.
Then there's the problem of power. As current handsets draw their power from the PBX, if the power goes out in an office, as long as there's power to the PBX then the phones are fine. IP switches don't natively supply power to devices plugged into them, so the options are to plug each handset into the local power circuit or purchase power over ethernet adaptors for each handset. Either way, adequate UPS backup is necessary to keep the phone system up in the event of localised power failure.
Regardless of the model you take though, there is still the inescapable fact that your voice needs are now tied to the network/Internet. If access to either one goes down, so too do voice communications.
What is convergence? In the IT world, convergence happens when two existing but different technologies or systems gradually draw together until they become one single integrated system. When talking about VoIP, convergence happens when your voice communications are handled entirely by the network infrastructure. Most businesses looking at migrating to VoIP will not be able to achieve true convergence straight away. This is largely due to the prohibitive cost of performing a complete migration from traditional telephony to VoIP. Assuming that the network infrastructure is capable as-is of supporting VoIP and requires no further improvement, there is still the cost of replacing every phone point with a network point, and replacing every non-VoIP handset with a VoIP-capable one. It's a lot of money all at once, and most businesses will find themselves in a position of gradual migration, supporting both VoIP and non-VoIP users. This requires a different implementation approach -- more on that later.
Additionally, it's important to remember that VoIP is, to an extent, still reliant on the traditional phone network to connect to non-VoIP systems. At some point, the call will have to route through the PSTN to reach its destination, depending on who you're trying to contact and how.