March 2003

Profitable Migration To IP Centrex

BY BRIAN MAHONY

Five years ago, I was a bright-eyed marketing/BD/strategy guy at a time when the industry was hailing 1999 �The Year of IP Centrex.� The VoIP softswitch was only recently defined, but fantastic multi-billion dollar IPOs of companies like Sonus Networks had everyone convinced that the softswitch and its most prominent linked application, IP Centrex, would sweep the telecom world. What�s more, everyone expected the merits of softswitch technology to push the large incumbent carriers to embrace this disruptive technology overnight. With a central attack by heavily funded next-gen service providers like Level 3 and WorldCom, and flanked by the viral growth of CLECs and their nifty new softswitch-based broadband services, telecom giants like AT&T and the RBOCs would have no choice.

What happened?

You all know what happened. The telecom industry was focused on the dual extremes of financing and technology with only a tenuous link between the two. On one extreme, overfunded vendors and service providers alike were chasing too few customers with a real demand for VoIP-based services. Many of these companies were more interested in impressing their investors than their customers. On the other extreme, vendors and service providers were demonstrating fantastic things with VoIP technology in their labs, but in most cases there was no practical way to deploy the technology on a wide-scale. The effects of this colossal disconnect devastated the telecom industry (and many of our stock portfolios) and the vestiges of what now seems like ancient history lives with us today.

Phew! Now that I got that off my chest let�s talk about what this means to us survivors in the telecom industry. Was this all a stock-option induced, fever-pitched orgy of IPOs and corporate malfeasance? Was it just a dream?

The answer is no. Though the hang-over still lingers (stock portfolios lost 90 percent of their value; job security seems tenuous at both start-ups and big companies), I am beginning to see a big ray of hope that the VoIP softswitch and its IP Centrex offspring will finally live up to their destiny and help lead us survivors to the Promised Land. This can only be done the old-fashioned way -- with a healthy dose of both smart technology and good business sense. Let me explain.

Phase 1: Get Minutes on the VoIP Network

The primary objective for all service providers deploying softswitch-based VoIP services is to get minutes on the network. This will generate the revenues and economies of scale to allow the provider to deploy more sophisticated services later on. The biggest mistake many carriers and enterprises have made in deploying VoIP is that they take on 80 percent of the complexity and costs up-front, and enjoy only 20 percent of the revenue streams, instead of the other way around. This has led to profitless deployments and service providers and enterprise customers who are frustrated and disappointed with the results.

The reason 1999 and 2000 and 2001 and 2002 each were not the �Year of IP Centrex� is that based on the way the service was defined in the past it became seen as a toy by the carriers. It did not meet the needs for a majority of their customers. Eighty percent of carrier revenues and almost all of their profits come from mid-to-large enterprise customers. These customers have many different types of sites. They have large campus sites with hundreds or thousands of employees that already have a fully functional and fully depreciated TDM-based PBX. These sites are not good early targets for IP Centrex because the current economics are sound and the enterprise has the on-site IT resources to support these legacy switches. They also have mid-size sites with maybe a couple hundred of employees. These sites are the perfect target for IP-PBXs like Cisco Call Manager or Nortel Meridian. Cisco and Nortel alone now ship in excess of 600,000 IP phones per quarter, mainly targeting these mid-sized sites. Finally, enterprise customers also have hundreds even thousands of smaller sites ranging from 50 person offices, and remote call centers, to employees working from home. This is the sweet spot for IP Centrex services. But before we get into that, let�s talk about what the carrier should do for the mid to large sites.

In Phase 1, the best way to capture the traffic from 80 percent of enterprise customer sites is to deploy a multi-service VPN. The good news is that QoS enabled data VPNs are growing by leaps and bounds. Many enterprises have taken into their own hands the task of building a secure managed network (based on technologies such as MPLS) to accommodate the data traffic across their sites. Over the past two years, the trend has been to outsource this managed network to service providers such as Equant, AT&T, WorldCom, and Genuity. With the addition of a VoIP softswitch handling Class 4 call routing, dial-plans, provisioning, and billing, these enterprises can now enjoy the benefits of a multi-service VPN that now includes data, voice, and even video services.

The benefits include virtually free calls on the network (roughly half the phone bill for enterprises is employees talking to each other), much-reduced long distance and international calls (using the service provider�s global points-of-presence), and reduced IT costs due to the synergies of a converged network approach. For the service provider, the benefits include increased revenues, reduced customer churn because of the added differentiation, and an order-of-magnitude reduction in operations.

The other hidden benefit of this Phase 1 strategy is that the end-user never knows it happens. There is no need for new CPE equipment, and the end-user does not change how she interacts with the phone. The change can be seen based on the cost savings the enterprise CFO sees. But the change is also psychological, both the service provider and the enterprise experience and begin to understand the value of a converged network. This understanding, and the underlying softswitch technology, becomes a launching platform for more sophisticated services later on.

Phase 2: Integrate with Multi-Protocol Enterprise CPE

This may seem obvious, but it was and continues to be overlooked by vendors and service providers alike as a crucial requirement for deploying managed services. For years now, there has been this conflict between service provider-based and premise-based solutions. Let�s end this today. The answer is that both approaches are required to deploy wide-scale VoIP managed services. IP-PBX deployments are still growing, and legacy PBXs are all being upgraded with VoIP cards. Up to 95 percent of this premise VoIP technology, including Cisco and Nortel and Siemens, is based on the more mature H.323 technology, but still today few providers see the problem with a SIP-only approach to IP Centrex services (let�s not even get into some limitations of SIP today, such as a standard way to exchange DTMF). The answer needs to include a multi-protocol, multi-vendor, hybrid network/premise approach. This means a multi-protocol softswitch that supports not only SIP, but also MGCP and H.323 protocols. In this way, the service provider can offer a multi-service VPN service that has the ability to interconnect with premise-based PBXs (through gateways) or IP-PBXs.

Supporting multiple protocols has the added benefit of giving the enterprise customer a wider choice of IP phones, and since this is the greatest cost for an IP Centrex service, multi-vendor translates into less cost. If the customer is looking for a low-cost solution, they can use MGCP/L IP phones that are much easier to manage than H.323 and SIP phones, and much closer to traditional PBX stimulus protocol phones (providing expected services like off-hook message notification for example). A range of MGCP/L analogue adaptors exist that bring the cost per port of MGCP solutions below $50 (these phones benefit from the intense competition in the cable market which is MGCP based). SIP and H.323 phones can also be used, but these more intelligent phones are also harder to manage, and you will experience many more interworking problems between the network-based softswitch (controlled by you) and the premise-based IP phones (controlled by different CPE vendors).

Phase 3: Deploy IP Centrex Integrated With Other Applications

So, with the Phase 1 and 2 revenues and cost savings pouring in, and with the required paradigm shift about the value of VoIP services, we are ready to deploy IP Centrex services. The difference from yesteryear is that this IP Centrex deployment will not be disconnected from other managed services like the multi-service VPN. Instead, it will be seen as an outgrowth of it. Using the same management and provisioning and billing interfaces of the VPN, and achieving relative feature parity and transparency with the premise (IP)-PBXs, IP Centrex will extend the service to allow direct service provider control of IP phones sitting on the employees� desks. Rather than warring with PBXs, IP Centrex will be offered were it makes the best business sense -- remote sites, SOHO, call centers, international sites, sites without IT resources, new sites, sites with IP phones, sites with only analog phones (managed via a station-side gateway via MGCP or H.323), wherever. IP Centrex will offer a predictable flat monthly fee per user and will be managed pain-free for the enterprise. Users will have access to integration with their desktop environment (such as Microsoft Outlook) and will use a Web-based portal to control and customize their service (call forwarding rules, find-me/follow-me, call lists, and the like).

Further, this integrated telephony environment becomes a platform for the quick adoption of other managed services, including network contact centers, video conferencing, voice portals, directory services, etc. The hard work has already been done to deploy the softswitch and customize the provisioning and management interfaces, so the ROI of each new service can leverage these prior investments.

So to conclude then, we can embrace the new-found pragmatism of our industry and find profitable approaches to deploy managed services such as IP Centrex. Let�s not make the mistakes of the past by getting lost in religious battles about VoIP protocols or pitting the service provider against the vendor of his enterprise customers -- these approaches might keep the dream of IP Centrex alive, but they won�t make it a reality.

Brian Mahony is the director of marketing for NetCentrex. He is a frequent speaker and writer on next-generation telephony and convergence issues. For more information, visit the company�s Web site at www.netcentrex.net.

[ Return To The March 2003 Table Of Contents ]