SIP trunking pilots and rollouts: 6 steps to success

You’ve decided the benefits of SIP trunking are worth ripping out those old PRIs. You’ve chosen the best SIP trunking provider in Canada to make the transition. Now it’s go time — but without following some essential steps, for many IT departments it could wind up being more like “stop and go” time.

In a recent virtual conference on SIP trunking best practices produced by Enterprise Connect, a pair of U.S. consultants, Jim Allen and David Stein, walked through the five phases of a typical SIP trunking deployment. While I think many technology professionals are becoming acquainted with the early-stage aspects of SIP trunking, the tips they offered on pilot projects and full-scale rollout sounded more specific and useful, so I created the following tip sheet. Keep this handy as your SIP trunking deployment gets underway.

Think about your dial plan. Most companies with an IP-PBX in place will find that the move to SIP trunking opens new opportunities to rethink the way calls are routed to end devices. These dial plans could be specific to a single site or, more likely, a group of sites.

What’s old must become new again. Session border controllers become key to integrating with legacy A323 signalling environments, Allen said, or providing interoperability between disparate SIP trunking equipment manufacturers. “While SIP is a standard, there is a still interpretation that’s occurred during the course of time,” he said. The SBC also plays a strong role in IT security, providing a demarcation point between your firm and external sources similar to a firewall, Allen said.

Imagine what your CEO will want. Your Session Manager may be bundled with your SBC and can provide a more high dial management component, Allen said. In other words, as you start to aggregate multiple PBXes together, you can create routing policies that would put a priority on routing calls to senior executives, or screening calls where necessary.

Start with the “don’t call us, we’ll call you” approach. When you start porting inbound traffic, it can be much more difficult to back out later, Allen said. Instead, take 10 DIDs and test extensively on outbound calls. Document call flows and explore as many scenarios as you can. For example, you may run into regions or countries that handle SIP traffic differently. Test voice mail and auto-attendants.

Don’t risk annoying customers. Consider the contact center. Allen and Stein recommended using G.711 for this area, as lower codecs such as G.729 may lead to failures in voice recognition prompts. No one wants to shout into a phone to get through your SIP trunking environment.

Voice should never be a one-way affair: The consultants said one of the more useful tests could be agent transfers from IP to non-IP, to make sure voice carries through to both sides. Use pilot numbers as part of the testing phase and route calls to a specific agent or cue in order to keep a closer eye on things.

Get even more practical advice by downloading the SIP Trunking eBook: Expanded 2013 Edition, from Allstream.