After spending this fall at a number of industry events, talking to customers and giving some serious thought to the topic, it's now becoming clearer how SIP trunking will really play a role going forward.

First, a couple observations / predictions:

#1 - SIP Trunking is really a misnomer - it really should be "Media Gateway Service Provider".  Think about it - if you really distill it down, the concept of SIP trunking is just about where the media gateway resides and who owns it.  For the foreseeable future, the vast majority of calls from your home/office will terminate on a PSTN phone somewhere.  I'd guess that 98% or more calls still terminate on PSTN circuits.  This means that virtually all calls must go through a media gateway somewhere to get to the PSTN - so it's really about who owns the equipment.

#2 - SIP Trunking is about consolidation - much like many other business models, the opportunity is in consolidation of traffic, sharing common media gateway infrastructure across a number of subscribers.  So instead of each business owning their own media gateway and PSTN lines, if you can get businesses to share a larger media gateway, you make money while they save money. 

#3 - "Security" Dare I say it? - very few SIP trunking service providers have really nailed the security puzzle of SIP trunking correctly.  Most push the problem to someone else or depend on ineffective SIP firewall products.  IMHO - this is like TJ Max depending on WEP to secure their credit card transactions -  a corporate disaster waiting to happen.  Any CIO worth their salt will make sure to have a Session Border Controller (SBC) protecting their organization from outside attack.

#4 - Tragedy of the Commons - by building a business model that depends on consolidating customer's traffic together through a common resource, someone eventually pulls out their Erlang calculator to figure out the usage factor and number of ports available for all the subscribers/customers.  This may be fine for some customers, but most of the customers that are really enticed by cost reductions don't want or can't share resources with others.  Think call centers or financial services - applications that push thousands of calls per hour into the network. 

#5 - The numbers may not add up - for many organizations, the cost of securing their SIP trunks and the service levels required of their SIP trunking service provider will wipe out any potential savings.  I've sat with customers and helped them do the math - it doesn't work in many cases.  However, for others, it does work - their needs are different and the model does work, even with the costs of security and dedicated bandwidth.

#6 - SIP Trunking is just a stepping stone - In the end, peer-to-peer SIP is where we are all headed.  We don't need a middle man to deliver our email, why do we need one for real-time voice and video conversations? I can easily envision enterprises using similar techniques we use today with email to call from enterprise to enterprise - all via SIP.  In the end, SIP trunking will be a transitional technology to reach those on "the old PSTN".  Think SkypeOut. 

I'd love to hear your comments, feel free to post them.