SIP Trunks and Contact Centers
Enterprises are starting to migrate, but we're just at the beginning.
Our most recent Enterprise Connect Webinar dealt with an area long seen as a potential quick win for SIP Trunking: The contact center.
The reasons are pretty straightforward: The classic SIP Trunking implementation features centralized communications servers using a single SIP Trunk (or mirrored for redundancy) to aggregate all enterprise traffic bound from and to the public network. The potential savings come from more efficient traffic engineering and bandwidth savings as well as potentially better rates from the provider.
The contact center is a natural place to test out SIP trunks because it may be easier to centralize a subset of your communications infrastructure--i.e., your ACD/contact center infrastructure--than it is to try and move the whole system over, at a time when you're likely still sweating legacy PBX assets across many sites. Also, since most contact centers are inherently B2C in nature, this represents a large chunk of your enterprise's traffic going to and from the public network.
Our webinar featured Art Schoeller, a principal analyst at Forrester Research, who offered the following recommendations for companies considering SIP Trunks for their contact centers:
* Be an "early pragmatist"--do a Proof Of Concept now
* Benefits exist for not only networked contact centers, but single site as well
* The proof of concept project should be real enough to prove in the economics
* Carefully assess contact center applications, IVR/Speech, Predictive dialing, DTMF signaling for takeback and transfer
* Plan extra time since trials are taking longer than planned
* Your incumbent carrier MAY not be as responsive as you like
* Deployment partner must have experience with carrier, SBC, and the rest of the ACD/session manager environment
Art also presented a chart with some Forrester data on SIP Trunking deployment that I found really interesting:
I think this is the first data I've seen that quantifies how much SIP Trunking is replacing PRIs, and as you can see from the bottom bar, there appears to be significant progress already, with 37% of respondents saying they've implemented SIP Trunks as a PRI replacement.
But it's going to be very much a staged transition. The next tier--those who are interested but haven't implemented--gets the market over the 50% mark when they do finally replace their PRIs with SIP Trunks. Then, presumably, the 19% who are interested but don't have plans will, over this time period, start developing their plans.
The last to go will be the 20% who are "not interested" today. Presumably at some point, as happens with all carrier WAN technologies, in a few years the service providers will start phasing out their TDM infrastructures for access, and PRI customers will then be pushed over onto the IP technology, ready or not.
The session also featured caveats about doing contact centers with SIP Trunking. I thought one of the most important was Art Schoeller's suggestion that enterprise managers have to be aware what sort of caller information/data is being passed in-band with PRIs today, and make sure that this information is being captured and transmitted in the IP format.
Our other speaker on the Webinar was Kris Hopkins, VP of Strategic Product Management at Acme Packet, who ran through some numbers that suggest the payback period for consolidating your ACD/contact center infrastructure could be as short as three months. Kris also emphasized the potential savings to be garnered from consolidating your call recording infrastructure, which can cut down on your need for servers by 50% in this area, he said.
You can register and see a replay of the webinar by going here.