Melissa Swartz | No Jitter | August 6, 2013
Voice services are much more complex than they appear, and changing out a telephone system is no exception.
Voice services have, historically, had a high level of reliability. This has led many who are not familiar with the technology to assume that because voice services are reliable, they must therefore be simple. This leads to a further assumption: Because it’s simple, anyone can do it.
Well, maybe not.
Voice services are much more complex than they appear, and changing out a telephone system is no exception. As consultants, we have been involved in the replacement of hundreds of systems over the years.
Here are the 10 most common mistakes we have seen when replacing a telephony system:
1. Needs assessment? What needs assessment? In many organizations, the goal of replacing the telephone system is simply to duplicate the existing system. However, telephony has changed considerably in the last 5 years, and new systems offer opportunities to improve communications and productivity in the organization. A thorough needs assessment, which covers both business and technical requirements, has multiple benefits:
a. Discovering productivity enhancements offered by mobility and collaboration capabilities not available in the current system.
b. Revealing communications issues. Often we find that many of the user complaints can be resolved through changes in business processes.
c. Improved system acceptance. When business users have had input into the process, they are less likely to resist adoption of the new system.
2. Assuming that all systems have feature parity. While this is true of basic features such as Hold, with newer capabilities (especially Presence and Mobility) there are significant variations among products in terms of how much information is available and how it is presented to a user. It’s crucial to see demonstrations of key features, and even better to involve business users who can provide feedback on which options work best for them.
3. Not involving business users in the decision. When a system is selected only by the IT team, the solution is often more technical and may not be as user-friendly. We have seen decisions that were based on what would be best for enhancing the IT team’s resumes. Involving business users in the decision process results in a better decision in the end, although a larger group can slow the process. Business users often help to clarify what is truly needed and what is nice to have; they typically aren’t swayed by the coolest new toys, and they don’t pick a solution based on how it will impact their resume.
4. One-Size-Fits-All Decisions. Most telephony solutions support a wide array of endpoints (big phones, small phones, softphones, wireless phones) and capabilities (Presence, Mobility, Collaboration). The variety can be overwhelming, so the options often get narrowed for manageability. However, most organizations have some percentage of power users who will benefit from higher-level “tools”. The most successful implementations recognize these users and provide the tools to make them more effective.
5. Skimping on user training. In many proposals we have reviewed recently, we have noticed a trend away from “classroom” training on live phones and toward webinars and train-the-trainer options. While these methods save a lot of labor for the vendors, they are often insufficient for the users. On the surface it seems OK; after all, how hard can it be to use a new phone? But new telephony solutions bring many new capabilities that users may have never been exposed to previously, such as the ability to manage calls and features using a computer instead of the hard phone; mobility applications; collaboration tools; presence; and more. In order to get full use of the new capabilities, users must be trained on them.
6. Overlooking analog requirements. While their numbers are dwindling, analog devices are not yet extinct. For example, faxes still have issues on an IP network. Most enterprises significantly underestimate the number of analog devices they are supporting. In some cases (such as alarms, equipment monitoring, etc.) the devices require a TDM analog signal and simply will not work if converted to IP. While such devices can be supported by analog lines outside of the enterprise phone system, this solution is not practical when there are many devices. Identifying these requirements up front may lead to selection of a hybrid TDM/IP system instead of a pure IP solution.
7. Ignoring the underlying infrastructure issues. This is really basic, but it happens all the time. Almost every new telephony solution has a significant VoIP component, and this means that the data switches supporting the IP or SIP phones must support VLANs and QoS.
Furthermore, these capabilities must be configured and working on the network. If you want your phones to work during a power outage, your switches must also support Power Over Ethernet (POE) AND you must have sufficient UPS resources in your data closets to support the phones. Finally, your cabling must be at least Cat5e and phones must be located within 328 cable feet of the supporting switch. When a phone is not located next to a computer, often there is a cable or distance issue. Sometimes meeting these requirements can be quite costly, and care must be taken up front to assess the environment and determine how these factors will impact the overall cost of the project.
8. Skipping the network assessment. Many people say “Our network will be able to support voice without any problem.” But voice is more than “just another app.” It’s an application that requires real time delivery of packets; they cannot be delayed or re-sent and still make sense. A network assessment will determine if voice traffic will be impacted by any jitter, delay or latency issues that don’t impact data traffic. Having such an assessment assures that the issues can be resolved BEFORE the voice traffic is impacted. The assessments do take some effort up front, but not as much work as diagnosing and fixing a problem after the fact.
9. Discounting 911 requirements. While not all states have e911 regulations that require the identification of a caller’s location within a building, it is necessary to properly identify the building where a 911 caller is located. Enterprise environments with centralized trunking or multiple buildings can create problems. Remote and mobile users are also challenging. Requirements must be fully understood up front and communicated to the vendors so that proper capabilities can be built into the new telephony solution.
10. Not thinking of the long term. While the selection of the new solution is a big undertaking, once it’s complete, you will have to live with the selection for several years. It’s important to select the right partner for the journey, and it’s really a two-part decision. Part 1 is the selection of the actual system/solution. Part 2 is the selection of the VAR/Integrator to install and support the system. The company you choose to support the system must be a capable partner who is experienced with the selected solution. Go beyond checking references; ask about the number of support personnel and their certifications. It can be very helpful to visit a vendor’s support center to get a better idea of the scope of their operation. You don’t want to be stuck without support if a key engineer leaves the company.
Which system is the best? Of course, the answer is: It depends. Here are a few of the factors you should weigh in making the judgment.
Make sure that everything you wanted in the contract actually makes it in, and that acceptance and payment terms are reasonable.
The process for porting DID numbers should be very predictable and well defined, but often it is far more difficult than it should be.
Nevertheless, the inconvenience today may well yield better results at the end of the process.