Over the past decade, the phone system business has changed dramatically. It changed from TDM to VoIP, from hardware to software, and from switching to unified communications (UC). The changes are not slowing. Traditional phone systems are threatened by cellular carriers, SaaS solutions, and non-traditional alternatives such as Skype and other peer-to-peer offerings as well as enterprise server based solutions.
User requirements and expectations continue to increase. Enterprise customers demand more features, more integration, more flexibility, and more manageability; all at lower costs with a quicker ROI. The voice component of overall communications may be shrinking, but quality voice systems remain a critical component to the enterprise comms puzzle.
The traditional switch vendors have largely developed server based applications to enhance their core switching platforms with modern UC capabilities. A similar pattern emerged among the players; multiple servers for each application consolidating into fewer servers with an increasing independence from the core switching functions. This independence is possible as a result of separate platforms, the increasing popularity of SIP, and the need to integrate with multiple systems and serve multiple locations.
Many of the switch makers are now positioning their UC products as a value-add to existing "legacy" systems rather than replacements. This makes sense given the current economic climate, and generally centers around the virtual number feature. This model effectively separates key PBX/UC intelligence (AKA applications) from the switching core. Drawing a boundary between the phone system and advanced phone features allows users to consider a variety of solutions rather than limit themselves to the incumbent switch maker or to any one PBX manufacturer.
It is a very interesting concept, and one that can apply to green field implementations as well. A brand new cutting edge UC solution could make sense to start with a bare bones on-premise PBX. A bare bones PBX is defined to possess the following core phone features: hold, transfer, redial, speaker, speed dial, and name/number display. The switch itself must support common features such as intercom, paging, PRI, T1, and ideally SIP trunking. No advanced features such as ACD, voice mail, or even redundancy. Most vendors offer such a product either as a separate model or as a bare-bone version of their primary platforms with minimal licensing. These base systems are relative bargains compared to their full-featured brethren.
It is worth noting that this bare-bone PBX could even utilize digital phones. It is a common misconception that VoIP phones are required for advanced features--most “VoIP” features are implemented on the switch itself or the desktop computer; not the phone. Other than the micro-browser, modern IP phones really don’t offer any additional functionality over their predecessors of the past 20 years--and few users actually utilize the micro-browser. VoIP phones are attractive since they enable a single network for management and reduce costs associated with moves. Digital phones should not be ignored though--they cost less and can utilize existing wiring and simplify LAN design.
The typical user will now have 2-5 private numbers such as their direct dial number on the bare-bone system, their mobile number, and perhaps their home number. The virtual number feature provides one public number that routes incoming calls based on user defined criteria--all or some private phones could ring depending on the rules. This is one reason why redundancy wasn’t important on the bare-bone system (it is important on the virtual number server), as calls can be routed to different locations should a problem arise. Call routing rules can determine routing based on time of day or even by who is calling (call screening).
The virtual number solution works very nicely for incoming calls, but outgoing calls pose a problem. The issue is the model falls apart when the private numbers become public; typically via outbound caller ID. Once people start directly dialing the private numbers, the value of the virtual number solution is lost. Multiple work-arounds exist, but the easiest/seamless solution is to just be able to pick up the (office) phone and dial. This can be accomplished with equipment and carrier choices that allow caller ID substitution (fairly common feature set with T1, PRI, and SIP trunking). Outbound calls from the bare bones system should send out the public direct dial number.
But the stated goal was a cutting edge UC solution. That definition requires some or all (even most?) of the following components:
* Presence
* Unified Messaging
* SIP
* Mobility component
* Fixed Mobile Convergence
* Instant Messaging
* Conferencing (voice and/or video)
* Collaboration
All of these components are readily available from one or multiple vendors. The bare-bone PBX provides core infrastructure including physical phones. The UC components can now be added. A UC solution including the virtual number service can be obtained/implemented in two basic ways: purchase the solution or as a service subscription. There are combinations and permutations of these two methods. An owned solution (from one or multiple vendors) can be implemented on-site, in a co-lo facility, or even in the cloud. The service can be purchased from a single provider or numerous specialized providers. Or, it can be a mixture of some on-premise and service based solutions such as owned/managed Microsoft OCS servers integrating with a SIP/UC service.
On the ownership model, multiple vendors offer this exact solution. The Avaya Aura Session Manager provides a UC umbrella for PBX systems regardless of vendor. Mitel’s Series-X service positions UC services for “legacy rescue”. Aastra positions Clearspan as an “overlay” product. There are many more, too many to name. The point is the vendors are now embracing dumb phone systems as components instead of insisting on replacement for their UC solutions.
Many of the manufacturers are working toward a software only version of their UC solutions. This is part of the general migration away from hardware, but also importantly positions their products for cloud-based deployments. But even as hardware today, the solutions lend themselves to offsite deployments. There are many factors to consider in location--security, disaster recovery, facilities management, location of users, and location of other servers. This final point is important since UC voice services are now and will continue to be integrated into numerous business processes and applications.
UC is much more a concept than a product or feature. When evaluating UC solutions, a user should not select a vendor based on specific features. UC is a foundation to build upon, and its architecture is what will determine its abilities around integration and to support web communities resulting with specific capabilities.
A UC service in front of the bare-bone PBX creates an interesting and strong value proposition. Service providers are rapidly adding virtual number services surrounded by numerous advanced features and UC applications. Phone.com announced a new offering with published pricing from $4.88-$18.88 per number per month (bulk pricing presumably lower). This service includes voice mail, vmail to email, auto attendant, advanced call handling rules, click to dial, call transfer (transfer from your cell phone), personal portal with call logs, and more. Factor in the bare bone box, and the resultant feature set and costs become pretty compelling.
Google is also entering this fray with Google Voice. Though their enterprise strategy remains unknown (and unavailable) it is likely they will offer a fee based enterprise version of Google Voice with number portability and DID ranges. The consumer version, expected to be released soon, offers a number of unique features such as voice mail transcription, call recording, and conferencing services. Google Voice and Gmail share Google’s Contacts, and presumably Google Chat (IM, Presence, and video) will be integrated soon.
A single service provider may not have all the answers. Call centers require specialized services that a general service provider may not. On-state.com offers a UC portal approach to call center management that factors in skills, presence, and communication method (phone, Skype, email, IM, etc.) into a single call center solution. Perhaps a different service provider is more appropriate for collaboration or international services. Yes, too many vendors can be a problem, but the traditional all-in-one PBX solution can be too limiting.
Presence/IM can be addressed for free from a variety of services or software products. Skype has received considerable attention lately. Integrating/managing Skype numbers into an enterprise dial-plan isn’t for the faint of heart. Some services now make Skype integration easier. Gizmo can integrate Skype with numerous services including Google Voice. A Google Voice call can be routed via a Gizmo SIP trunk resulting in virtual number service without any Skype usage charges. Add some of these services and the bare-bones PBX feature set expands to include presence, collaboration, and videoconferencing. Factor in the virtual number service and the user gets fixed mobile convergence, conferencing, and unified messaging.
For customers that want to own the UC solution rather than subscribe to it, the cloud represents an emerging model of interest. Cloud computing turns the hardware into a scalable service. A UC product could be implemented on "hardware" in a cloud service such as Amazon's EC2. This offers some of the benefit associated with the service model (outsourced hardware), but offers total software control (upgrades, customizations, and APIs) to the customer.
This is where this article transitions to science fiction. There is very little telecom/UC solutions actually actively being marketed for cloud computing at this time. There are a few players well positioned for it, but most of the traditional players are still working to untangle their hardware from their software solutions.
The vendors best positioned for cloud computing are typically the nontraditional players coming from a perspective of hardware independence. Siemens took the software-only plunge fairly early among the traditional players and as a result were able to demonstrate a "Proof of Concept" at the recent VoiceCon show in Orlando (video here). Their demonstration took the traditional set-up configuration pages into an Amazon storefront--blurring the distinction of customer owned software with a cloud and web based service. For example, in their demonstration, the set up screens allowed the customer to order licenses and purchase phones utilizing Amazon’s storefront and Amazon’s cloud servers.
Another vendor well positioned for the cloud is Aastra. Currently, Aastra is selling their ClearSpan UC/PBX coupled with blade server hardware. ClearSpan is based on Broadsoft software which is a popular platform for IP Centrex providers. That is a lot of service model experience already built into a software license. Microsoft will also fare well with the cloud now that providers such as Amazon support Windows environments. OCS, Exchange, and SIP for services and carriers could all add-up to a strong cloud offering, particularly when complemented with a bare-bone PBX.
All this UC innovation ironically results in strong reasons to consider a bare-bone phone system. The core systems are very good at basic things; such as 911 and intercom. The base phones meet most user needs, and the costs of these base systems can be a third of their feature rich alternatives. Combining the bare-bones system with one or more services or products for UC capabilities and virtual numbers can result in a very powerful solution.
This article was written by contributing author Dave Michels. Dave is a principal at Verge1 Consulting and blogs regularly at Pin Drop Soup.com.
