telecom consulting service providersExplore the Hosted VoIP vs. IP PBX debate in our blog. In this exploration of communication technologies, we aim to shed light on the debate between these two prominent options. Join us as we navigate the intricacies of Hosted VoIP vs. IP PBX, examining their features, advantages, and considerations for businesses seeking the most effective and efficient telecommunication solutions.

Hosted VoIP vs. IP PBX Debate

The following is a discussion amongst telecom consultants over which business phone system solution would work best, hosted VoIP or a premise based IP PBX:

Question: From a best practices standpoint – what are the pros and cons of a pure hosted Voice over IP provider versus using an IP PBX phone system equipped with SIP trunks?

  • Assume a setup of a company with 40+ domestic sites with 30 end sets at each site – for a total of 1500 devices or stations, all connected via a robust MPLS network.   If using Hosted VoIP, the provider has two redundant nodes on the network and the calls are transmitted into the MPLS cloud and terminate to SIP devices.  If using an IP phone system, there would be two different systems for redundancy, installed in different data centers (using your favorite brand system – Cisco, Avaya, Mitel…) also connected to the MPLS network.  The IP system has SIP trunks coming into it that carry the numbers for all the 40+ sites.  The VoIP provider has a nice customer portal to make adds, moves and changes.  The IP phone system does as well.
  • Assume also that the upkeep (hardware & software) is performed by the carrier and a 3rd party, in the case of the IP phone system.  The Hosted VoIP provider maintains their hardware and software.  Likewise, the IP PBX phone system hardware and software is kept up-to-date by a 3rd party value added reseller (VAR).
  • Assume – company does not need a robust feature set – smaller call center and help desk, conferencing, etc.
  • Assume – cost/pricing is the same – short & long run.  (not a reality but pertinent to the question).

In the short run the Hosted VoIP solution is simple, easy to deploy and gets the IT group out of the phone hardware business.   However, what are pitfalls in going this direction – short term / long term?

Despite your assumption that “cost/pricing is the same – short & long run”, cost actually IS the major issue. With hosted VOIP, you’ll need far more bandwidth, and that comes with additional recurring monthly cost. Quality hosted VoIP requires G.711 which takes up 90 to 100K per conversation and can fit up to fifteen simultaneous conversations on a single T-1 (I have yet to see consistent quality run over G.729, which only takes up around 30K per conversation).

With an in-house IP PBX, you don’t have that extra cost. Also, you won’t pay “per seat” as you would with hosted. I find 23 seats per site to be my general limit for hosted VOIP. Above that, it’s less expensive overall to get an in-house IP PBX.

For 40+ sites with 30 devices at each, an MPLS in place and redundancy necessary, I highly recommend Shoretel, especially for redundancy (as well as their call center module). Most manufacturers require “N x 2” architecture for redundancy, meaning that you have to have (and invest in) redundant systems at each site. Shoretel uses “N + 1” architecture, meaning that if one site’s PBX or outside lines goes down, the other sites’ PBXs seamlessly pick up the slack via the MPLS network and nobody notices anything except the admin. As far as their call center application, reports are customizable by the call center admin via a graphical user interface.

You can always outsource the basic management tasks on the MPLS and ShorTel (or others) for far less than a hosted solution. This would keep IT out of the day-to-day management of the system while providing more flexibility at a lower cost.   I would also make sure that my design did not include any locations with two different local loops, which will drive your costs up.  If capital expense is an issue you can buy the equipment from one of the carriers and they will turn the cost into an operating expense with very little persuasion.  Depending on the size of your data traffic, I would consider implementing WAN Acceleration, especially if it will give you a further reduction in your access optimization.  I would recommend Riverbed and you should see a 60%+ reduction in bandwidth requirements for you data traffic with the added benefit of increased response time by a factor of around three.

You have gotten some good feedback, but I’ll add a few things.

1.       In this situation, I think your sites are too big (30 seats) for hosted PBX to be economical, especially if you’re not doing any call center functionality.  Hosted systems are almost unparalleled in features and functionality and if you did a true apples to apples, a premise solution would be more expensive – but that would only be needed if your users were REALLY in need of all the functionality.   Based on your scenario, you would not need the call center, call routing, IVR, etc.

2.       When choosing hosted vs. premise you need to look at the life expectancy (or at least how long the business will depreciate the capital investment), and compare hosted vs. premise based on that many years.  You will find that hosted will look great the first few years, but there will be a tipping point where that hardware will cost you less over time.

My personal opinion on this is you will probably find that a centralized premise solution that is served with SIP trunks from a carrier that can handle your E-911 needs and a big enough SIP footprint to cover all of your phone numbers; and a hardware vendor to manage and create a quasi hosted environment – where you own the hardware – will be your best solution when considering costs, diversity, and not an over purchase of features, that in the end will never be used.

I would have to respectfully disagree. Given the requirements that have been laid out with less than complex needs, I would lean toward hosted and here is why:

1.       In a Hosted model, there is no need for MPLS for voice routing. Not only does this greatly reduce cost, it also takes a great deal of complexity out of the picture. Each site can use a managed circuit from the hosted provider (which is in essence a point to point to the hosted provider’s data center) and a cheap broadband circuit for data. Voice can fail over to the data circuit if the voice T-1 goes down.

2.       Time and time again I see PBX providers going end of life on hardware and forcing customers into heavy Cap-Ex upgrades. In a hosted model, features and upgrades are automatic and pushed out to the end user as they become available. Case in point: I had a multi-location company that we sold an Avaya PBX to in 2008. They are now ready to deploy to a couple of other locations. However, the equipment that they have (phones and software) is not available anymore. So, now only four years later, we have to change the user experience again and scrap 4-year-old equipment.

3.       Redundancy: As mentioned above, you don’t need MPLS, just broadband (which voice can fail over to). Beyond that, the top hosted providers are in multiple data centers riding redundant power grids and utilizing multiple carriers and transport methods. If your HQ is down from a hurricane, etc. (such as I’m in the middle of right now), users can simply log in wherever an internet connection is available an make and receive calls without the incoming caller ever knowing there is a problem. In a PBX world, you would be forced to build a private cloud in order to compete with that type of redundancy.

4.       There are huge soft dollars saved when you consider the fact that your IT staff can now fully focus on IT and not telecom.

5.       My best hosted providers are using G726 (30k per call) with extreme success.

I will say that hosted is NOT always the best option. However, in this scenario, I think it warrants a very close look. In response to the question regarding pitfalls, I think the pitfalls are more along the lines of disaster recovery and business continuation rather than cost. People can argue that the ongoing cost in a 10-year total cost of ownership comparison will sway the pendulum toward a PBX model. However, when you consider, network savings, increased productivity in your IT staff, zero PBX maintenance dollars, and no need for expensive hardware upgrades in that time frame, cost becomes much less of a viable argument.

By the way, perhaps even ShoreTel sees the writing on the wall, because they purchased M5 Networks in a move toward becoming a true cloud provider.

One item you should consider as well is Least Cost Routing option where you can connect to various SIP service providers. Use the least expensive route for your calls.  In my experience, going on 40 years in this business, it is the most cost savings feature of SIP services.  You will have to deal with 503s SIP error response codes, however, once you fine tune your routes, you will find that it was well worth your time.

If you would like to receive help implementing a new business phone system or any other telecom consulting service, please complete the web form on the right side of this page.

Call us at


for immediate service or fill out the
form and we’ll be in touch right away.

  • Please describe how we can help:
    What services are you considering to drive your business higher?
    What challenges are you currently facing?