Buying an IP-PBX: How to Make the Right Strategic Decision

The article that follows is based on John Riddell’s half-day seminar on preparing for the acquisition of an IP-based PBX, given on October 17, 2005, at the Telemanagement Live conference in Toronto. Buying a PBX has always been a difficult process, with many traps and hazards for the decision maker. But acquiring an IP-based system requires more preparation, attention from more corporate departments, and commitment to a longer-term perspective with the chosen vendor. It’s a strategic decision that demands intensive preparation. Pit Stop for PBX IndustryThe new IP-based systems marketed since the start of the decade turned out to be more complex and prone to trouble than the TDM (circuit-switched) systems they were replacing. Not only can the IP-PBX be brought down by perils new to telephony, such as denial-of-service attacks or viruses, but maintaining voice quality itself requires careful engineering. No wonder 63% of the 177 respondents polled by Empirix last year said they were "concerned about putting a VoIP phone on their CEO’s desk." Moreover, some of the early claims about the benefits of the new technology proved to be overblown. Several IP-PBX vendors have posted Return on Investment (ROI) calculators, which attempt to calculate the cost reductions and productivity gains available through IP technology. But productivity claims are hard to quantify and often overblown. For example, a Sage Research study of April 2005, indicated that IP telephony will save the equivalent of no less than 6.2 hours a week per employee. If this were true, investment in IP telephony would be fully paid for in one to two months, and almost all organizations would have adopted the new technology long ago. But in fact, the advent of IP-based systems has seen not a speeding up but a slowing down of the PBX replacement cycle. According to Infotech, enterprise lines shipped per year dropped 20% from their 1999 peak and are not yet up to their trend line. These market results tend to bear out the judgment of U.S. PBX analyst Allan Sulkin, who told the February 2005 VoiceCon conference, "No guaranteed cost savings can be attributed to IP telephony; and gains in productivity are difficult to assess and quantify." Or, as analyst Rohan Mahy asked the Next Generation Networks conference in September, "Name one application shipping on an IP phone not already available on a TDM cellphone." There are cer-tainly cost savings associated with IP, Mahy continued, but for most customers "they are minor when associated training costs and data network upgrades are taken into account." The PBX industry has in fact made a pit stop of several years duration, during which its focus has been on perfecting the new technology, and the pace of delivering new functionality to customers has slowed. But these are mere growing pains. IP Telephony TriumphantDespite its early problems, IP telephony is today firmly established as the reigning technology of enterprise communications for the foreseeable future. All PBX vendors (plus of course data-centric new entrants like Cisco) are carrying out all of their research and development in the framework of IP-based technology. Some of the innovations can be utilized with TDM equipment, but many cannot, and over time the gap will grow. Most PBXs sold today are IP-capable. The technology is now proven; applications development has speeded up; prices have fallen. We have now reached a turning point in IP-PBX adoption: the majority of lines shipped in the U.S. last year were IP, and Canada will likely pass this milestone in 2006. Current market projections are that by 2009, IP will account for 90% of lines shipped. TDM phones will continue to be sold and used for some time, but we can now say with confidence of almost every Canadian organization: "The next PBX you buy will be IP." For the enterprise decision maker, this is a formidable challenge. As Claude Lessard of McGill University Health Centre recently explained, "You must start to prepare.… It will be painful—but the pain will be small if it’s spread over time" (see box). I’m going to focus on four aspects of this preparation: > Choosing the vendor. > Upgrading the data network. > Preparing the staff. > Identifying needed functionality. Time to Change Vendors?During the 1990s, PBX acquisition became relatively routine. Most corporate users bought a new box from their established supplier every decade or so, preserving many components and sets from the old system. For most, switching vendors was only posed if the experience with the previous system had been negative or in cases of major corporate restructuring. By contrast, the switch to IP technology is a logical time to move to a new vendor. Consider: > Because the changeover is so comprehensive, the advantage of sticking with the incumbent, while still real, may not be as great. All things considered, more system elements need to be changed out. > IP has "decommoditized" the PBX industry. There is more differentiation among vendors in architecture, functionality, and customer support. > The transition to IP is arduous, even if you stay with the same vendor, once again reducing incumbent advantage. > Once the IP-PBX is acquired, incumbent advantage comes back with a vengeance. Changing vendors down the road will not be as easy. The last of these points is related to a change in the PBX life cycle. Circuit-switched PBXs were generally rated as having a life expectancy of seven to 10 years (although some lasted much longer). The sets and some other components lasted a decade or more. With an IP-PBX, some components have a shorter life cycle: the servers, switches, and routers must be replaced every three-five years. The sets, which are much more complex, will not last as long as previously. But most of the IP-PBX’s value resides in the software. And provided that the vendor continues to flourish, properly updated software lasts indefinitely. The investment represented by that software will be a reason not to change to a new IP-PBX vendor. Reshaping Corporate NetworkingAn IP-PBX purchase, over time, will have an impact on the organization extending far beyond pure telephony. To begin with, running voice over the data network changes data communications forever. Major upgrades to LANs and the WAN are usually required. Voice traffic also places constraints on the future evolution of data traffic: what will be the impact of bandwidth-hungry new data applications on voice traffic? On the bright side, state-of-the-art communications systems can often spark a transformation in significant businesses practices. The first step to success in an IP-PBX acquisition is therefore to discard the notion of a PBX-for-PBX replacement. An IP-PBX is new—in its requirements, its characteristics, its potential. Upgrading the Data NetworkYou’ve probably heard it said that every dollar spent on IP-based telephony systems generates on average three dollars in sales of data network upgrades. I recently asked a Cisco executive if this guess was close to the mark. "We don’t use that figure, but it’s not far off," he said. "I can say that when a company goes to IP telephony, they must replace on average 20%-30% of their data infrastructure." Other estimates support this view: > "Vendors indicate that the network is insufficient in 80%-90% of projects."–PlanNet, February 2005 > "85% of data networks aren’t ready for VoIP."–attributed to Gartner, 2005 Whatever the cost of data upgrading, it complicates formulating the IP-PBX business case: how much of this expense should be assigned to the IP-PBX acquisition? Data equipment vendors have been improving their switches and routers to build in capacities needed to support packetized voice. As voice-ready gear works its way through the market, the average cost of upgrading is declining. But for most organizations, the need to upgrade is still there. So too is an array of now familiar engineering, configuration and management problems, including: > Quality of service: Will high-quality voice transmission be assured by building in Quality of Service priority for voice packets, or will the user rely on overprovisioned bandwidth to do the job? > Voice compression: The many different codecs available to digitize voice have different performance characteristics. High-compression codecs save on bandwidth but their signal is more vulnerable to degradation. > LAN vs WAN transmission: Many organizations use higher levels of compression on their Wide Area Networks than in the local area. But passing a signal through multiple codecs can affect voice quality. > Power supply: IP phones draw about 30% more power than their TDM predecessors, and power must be provided by Ethernet switches. As a result, equipment closets may need an upgrade of power supply and air conditioning. > Emergency calling: IP phones can be moved around and plugged into any jack in any location—a big improvement, except that it frustrates locating emergency calls. The countermeasures are not simple. > Reliability: IP telephony provides new resources for achieving or surpassing 99.999% availability, but at a cost: redundancy must be built into not just the call control unit but the data network as a whole. > Security: IP-PBXs provide new vectors for hacker, virus, and toll fraud attacks, and no fix is yet in place to counter the looming threat of Voice over IP spamming. > Day Two management: An IP-PBX that works perfectly the day after cutover may fall victim to subsequent changes in the data environment. A flourishing industry of network monitoring providers has sprung up to meet this need. (For more on these topics, see Teleman-agement #196, 198, 201, 210, and 211.) Network AssessmentIt is only prudent, before signing to buy an IP-PBX, to arrange for a thorough assessment of the data network in order to identify upgrade needs. Software is available to do this task, but vendors ask customers to pay for this study, especially when no PBX purchase contract has been signed. The vendors’ stand is logical: their products are dependent on the health of your data network, which is beyond their control. From the customer point of view, this extra charge is irritating: it’s like being asked to pay to test drive a car. But inking the deal without an assessment can result in large, unexpected charges that blow the business case for the acquisition. Bargaining skills are needed here: if the vendor is convinced you mean business, they may waive the assessment charge. Prepare the NetworkersIt’s a truism of voice-data convergence that the first step is to converge the voice and data staffs. What’s surprising is that five years into the IP-PBX era, the difficulties of this challenge are still getting a lot of discussion, even among the most tech-savvy and proactive user organizations. For example, at last February’s VoiceCon conference, Craig Hinkley of Bank of America recalled early discussions of IP telephony in these terms: "All the engineers look at each other. The voice guys take a step to the left; the data guys take a step to the right. The trenches go up, and the grenades start to fly." Changing the organizational chart is obviously not enough. Many measures have been recommended, starting with crosstraining, but the end result must be that voice and data specialists assume joint responsibility for their converged environment. One of the early promises of IP telephony was that it would permit a downsizing of management staffs. Not much is heard of that today. Instead, Cisco Systems tells us that following their own IP telephony implementation, "Our staff needs didn’t dwindle through convergence. It brings on new capabilities in call processing, unified messaging, contact centre, Web collaboration, and videoconferencing." (Graham Hosie, quoted in Packet, Q1 2005) The promise of staff reductions heard so often a few years ago played to the passion for downsizing sweeping executive suites in those days. Today, a counter-tendency is evident. As Henry Dortmans, president of Angus Dortmans, noted in our September issue, "After years of ‘doing more with less’ or ‘everything with nothing,’ many enterprise business executives are starting to see the need for stronger networking management. If that is true, it means we’ll have a much better chance to utilize the new capabilities of today’s communications technology." Second Generation IP-PBXsAccording to some vendors, these capabilities are today advancing into a "second generation." The first generation of IP telephony, we are told, was the period of proving the technology and targeting cost reduction. Savings were promised in the reduction of long distance and PSTN access charges, bandwidth conservation, reduction in expenses for moves, adds and changes, and support of remote workers. In the second generation, the focus shifts to "business values and productivity." There’s an element of marketing hype in this talk of "generations." But there’s some truth as well. IP telephony, combined with new wireless technology, today offers a wide range of interrelated applications that go beyond what was current on PBXs even five years ago. For example: > Unified messaging lets a user manage incoming emails and voice messages off a single visual queue. Available for many years, it is now winning wider adoption in IP-PBXs. > Find-me/follow-me services integrate the many communication devices available to today’s mobile user, permitting all to be accessed from a single phone number. > Instant Messaging is now available in enterprise-grade versions that provide the security and archiving required in most organizations. > Presence posts messages regarding the user’s availability or unavailability and can filter calls depending on source. > Instant conferencing permits audio or video conferencing to be set up without notice, by manipulating icons on a desktop. > Collaboration applications enhance conferencing through file sharing or whiteboarding. Prospective buyers need to look not only at what is available now but also at some of the projects now in their research and development phase. Among them: Voice over IP is not bound by the 64 Kbps transmission limit of circuit-switched telephony networks. Codecs now available deliver high-fidelity voice far beyond the limits of toll quality, with or without an increase in bandwidth. Voice over IP needs to escape from the island of the corporate wide area network: the current goal is secure, spam-free IP calling anywhere. As Session Initiation Protocol wins wider adoption, it should be possible for a user’s SIP URI (Universal Resource Indicator) to emerge as the primary business phone number, linked to all user devices. It will soon be possible to switch devices during a communication session, moving smoothly from Instant Messaging to a voice call or vice versa, or from a desk set to a cellphone. IP-PBX users may soon be able to log in at any IP phone anywhere, obtaining their local profile. Presence-enabled callback initiates a call or conference at the moment a user becomes available. Predictive presence gives detailed availability information based on the user’s calendar, such as, "She’s at lunch but will be back in about 10 minutes." Desk phones and cellphones can now function off a single number. Soon, Wi-Fi phones will offer cellular capability. Location-enabled Wi-Fi phones could answer a question like, "Tell me when the CEO leaves the conference room." It is an open question how quickly applications like this will win acceptance. But putting together what is available now with what may be soon, we have the outline of what is likely to become a new standard for business communications. If you’re considering an IP-PBX acquisition, review whether present or future applications are a good match for corporate needs. And be sure that a chosen IP-PBX supplier excels in developing high-touch multimedia applications. Take the Long ViewTo sum up, even if PBX replacement is not on the agenda this year, users need to begin active planning now. Among the top priorities: > Integrating the voice and data staffs. > Evolving the data network to meet future needs for IP telephony. > Studying how the new capabilities of real-time communications systems can be used to improve or revolutionize business processes. Claude Lessard put it aptly, "The pain will be small if it’s spread over time." Lessard’s Law"You may not need IP telephony now, but over time, you really don’t have any choice in the matter. Brace yourself: they lied to you. One network is not simpler or easier to manage than two. But you must start to prepare. Start now, slowly. It will be painful—but the pain will be small if it’s spread over time." Claude Lessard telecom coordinator, McGill University Health Centre

Buying an IP-PBX: How to Make the Right Strategic Decision

News | 01/01/2006 5:00 am EST

To keep reading, get a free trial.

Already a Subscriber?

Reuse & Permissions

Inside The Wire Report

Information

Hill Times Publishing

Get in Touch With Us