Calling All Centrex Users: Your Cloud Solution Is Dead! (Part 1)

Calling All Centrex Users: Your Cloud Solution Is Dead! (Part 1)

By Stephen Leaden August 28, 2017 6 Comments
Stephen Leaden PNG
Calling All Centrex Users: Your Cloud Solution Is Dead! (Part 1) by Stephen Leaden


Are you a Centrex user? Have you even heard of the term Centrex? For those of us who are new to the UC space in the last 15 years, you may or may not have heard of the word Centrex, or “Central Exchange,” a.k.a. the original “Cloud.” Centrex back in the 70s was the mainstream telephony solution for most larger enterprises. Centrex provided some interesting features that were really worthwhile at the time:

  • High availability was built-in to every Centrex, which resided in the Central Office (I was responsible for an 8,000-line Centrex at one time). Redundancy and automatic back-up to generator were part of the solution and uptime was extremely high at 99.999%
  • Extensions native to the Centrex were available just about anywhere, as long as you were connected via an Off Premise Extension (OPXs) and were willing to pay for the mileage charges associated with such (mileage-sensitive circuits are non-existent today, although we do have a client with OPXs as I write this), similar to remote workers in today’s environment
  • Integration to other Centrex's and premise PBXs were available via tie trunks (analog at the time, then eventually replaced by T-1 "channelized" circuits)
  • Centrex offered based features such as DID service, voice mail, call accounting, transfer, conference, and bridged line appearances - all team members who had bridged appearances could see who was on the line. "Sophisticated" features included "ring, no ring" on any given extension on any phone, user choice, and speed dial. Of course, there was no internal programming available back in the day for soft moves and changes - most everything required a change order, or MACD (Move, Add, Change, Delete)
  • Most Centrex’s provided some level of Call Center or ACD, Automatic Call Distributor with minimal features at best. They included call routing and reporting and little else outside those basics
  • Centrex offered minimal Capital Cost outlay (CAPEX), most for implementation and turn up, while all else, including Centrex extensions and features, as well as end points, were rented, or OPEX

Now since all TDM digital voice communications have been at end-of-support for over 60+ months, Centrex (TDM as well IP) and its ACD affiliate have slowly been eroding. We have one client with a 7,000-line Centrex and ACD utilized for close to 30 years with capacity issues and end-of-support. The vendor has recently advised the client that it cannot add any more Centrex lines, period. In my own experience of 30+ years in the industry, I have NEVER witnessed a Centrex at capacity.

Both TDM Centrex and IP Centrex and ACD require too much maintenance, physical space, physical interaction, and continued specialized training for a legacy technology on the part of the vendor. All parts have been refurbished for some time. Therein lie the reasons for Centrex retirement.

So What are the Risks?

The risks associated with staying with a grandfathered legacy technology such as Centrex are significant. They include, among others:

  • System End-of-Support and Spare Parts
    • Manufacturer announces End-of-Life of their telephony-based system (Centrex in this case)
    • Spare parts are refurbished and increasingly become less available, maintenance contract support and SLAs for mean-time-to repair over time become an issue, and costs may become a major issue
    • Notices from the providers give 12-24 months’ notice on end-of-support even to this day, meaning that any larger enterprise has little time to replace such a system, especially if 1,000+ end points and above (Microsoft recently announced 14 month end-of-support for 3rd party UM integration, for example)
  • Capacity Issues with Current Systems and Associated Costs
    • Systems at capacity require additional hardware investment for a temporary solution (on-premise or cloud) to extend the current Centrex out further, and some increases by as much as 40% of a new system, in my experience
    • Capacities are typically a BIG flag for not having a lot of time left for vendor support
    • Capacities and temporary add-ons create a more complex maintenance environment
    • Note that one can always create a premise "hybrid" solution with tie trunks between the Centrex and premise system
    • Newer systems added onto older system to temporarily solve capacity issues is not a good solution unless it is the going-forward replacement solution
  • No Investment by Telecom Manufacturers in Traditional TDM Centrex and Premise-Based Voice Systems
    • 60+ months
    • TDM, for all intents and purposes, has been dead for years now
    • Centrex users are now being given notice for end-of-support similarly
  • Consumer-Driven UC Capabilities Are Here In The Enterprise and NOT Available in legacy Centrex
    • Mobility/twinning, Unified Messaging, IM/chat, presence, LDAP and Active Directory integration, Collaboration, softphones, desktop audio conferencing, desktop video conferencing are driving the need to replace legacy Centrex. Even Caller ID is not available on some legacy Centrex systems, really!
  • Fewer Trained Technicians
    • Knowledge base "how to fix" has eroded significantly
    • Difficult to find vacation, sick replacements for staff with legacy expertise
  • Systems Age (we use 7+ years or more as a baseline)
    • Any Centrex or premise-based system requires either some type of an upgrade or replacement at 7+ years in place, and 10+ years is a clear indication of the deterioration of support for parts and technicians
    • Expect manufacturer notices, spare parts unavailability, upgrade unavailability, and fewer trained technicians over time
  • Risk of a Multi-Day Outage Increases
    • State of the systems poses greater risk to basic communications services - a single outage could last several days.
    • How would your organization communicate in the event of a multi-day outage? (Several financial enterprises encountered such after Hurricane Sandy in NY City metro area, now Hurricane Harvey in Texas)
    • What risk(s) would this pose to the safety of the staff and community?
  • "Good Money After Bad"
    • Investing in older technology with no going-forward vendor investment for expensing on capacity equals "good money after bad," and a high probability that the short-term investment will be retired for pennies on the dollar in the near future

Conclusion (Part 1)

Yes, there is risk with your Centrex environment, and something to seriously consider. Interestingly though, the "hiccup" in migrating from a legacy Centrex to a new UCaaS or CaaS solution may not be as far a reach as one would think. Both are OPEX models and the biggest hurdle financially when migrating to the cloud.

If you are already on Centrex, then you paying into an OPEX model. Your financial model migration may not be as difficult to migrate to as would a traditional CAPEX UC model.

In Part 2 we will cover what an enterprise can do and how you can be proactive in the planning and replacement process for your legacy Centrex or legacy ACD.


6 Responses to "Calling All Centrex Users: Your Cloud Solution Is Dead! (Part 1)" - Add Yours

Kevin Kieller 8/29/2017 7:55:10 AM

Very good points Steve. I look forward to part 2.

Telecom providers have loved Centrex because it provided incredibly high margins. I know of one very large carrier who was achieving 98% gross profit on Centrex services. That being said, the platform was incredibly "brittle" and no one wanted to add or change anything for fear it would break something that no one remembered how to fix. (Seems like an idea for a dystopian sci-fi novel where the only communication system is Centrex and no one knows how to fix it when it breaks down one day due to a solar flare.)

Greg Zweig 8/29/2017 10:15:07 AM

Steve - unfortunately, I think the many of the proof points and details in this article are not entirely accurate. There are certainly customers that are being asked to migrate off their Centrex deployment but this is a business decision made by a specific carrier, not a broad technology issue. GENBAND acquired the Nortel carrier business (DMS) in 2010 and I can assure you that we still support tens of millions of TDM lines for carriers across the globe. More importantly, just because a customer has an old Centrex phone doesn't mean the carrier has old infrastructure running it. We've spent hundreds of millions of dollars on R&D to provide migration paths for our platforms and designed them to be transparent to end users (if they don't care to upgrade).

The software we ship today (all IP - running on commercial off-the shelf hardware) will seamlessly support an old TDM phone. In addition, customers can mix and match them with new SIP phones, UC clients, mobile clients, etc. as well as add a new call center. That's no accident, service providers have long required that we deliver support for legacy infrastructure and the latest services on the same core platforms. We have millions of lines deployed in this manner. I'll be happy to step you thru the nuts and bolts of this.

Additionally, you make a number of references to the age of carrier equipment etc. It's true that some carriers have been slow to modernize but it's dangerous to generalize. The proof of the pudding is in the eating, I think we should compare the carriers annual outages to what your customer's experience with IP-PBXs or cloud providers. The carriers will best them by a mile.

Look, I'm not one to cling to the past as I helped found one of the first IP-PBXs companies and I'm a huge advocate for cloud services. However, that doesn't mean that customers need to rip everything out to move forward. Before you publish Part 2 I hope I ha
Steve Leaden 8/29/2017 3:38:48 PM

Thanks Kevin

I appreciate your input. Fortunately, my own experience has been good to great with Centrex, just not the full feature/functionality we have come to expect from today’s cloud solutions. Centrex was one of the few solutions in the day, and yes, almost dystopian in that respect. Well said, Kevin.

Thanks again, Steve
Steve Leaden 8/29/2017 4:10:47 PM

Thanks Greg for your thoughts.

Let me clarify some of your points.

My recent experience has been with a carrier who decided to no longer support their Centrex in terms of growth and support, and placed the client in an awkward position. Note that I am referring to Centrex, not UCaaS, the original TDM cloud solution, based on legacy AT&T #5 ESS and Nortel DMS technologies, not newer IP-based technologies.

National carriers have been slowly retiring Centrex in all forms for years now, being replaced by GENBAND IP architecture, Broadsoft, or from-the-ground-up IP architectures, in my experience.

It is welcoming to hear that GENBAND continues to support TDM customers around the globe. In my own experience Unless you can validate otherwise, all TDM digital end points have been discontinued at least 24-36 months now with no upgrade path for additional features and functions or software releases. And I would highly doubt that GENBAND supports customers on legacy DMS series switches at their site without some type of end-of-support notice, or DMS switch in the cloud powering TDM users.

A point was made referencing GENBAND’s investment in software-based IP architectures. From my viewpoint, that is UCaaS and not legacy Centrex.

Regarding age of equipment, I am making reference to circuit cards, both station and trunk side, digital end points, and other hardware-based platforms, most of which have not been supported for well over 60 months. We have a client that migrated from a hardware-based GENBAND DMS-100 to a CS-2100 and the square footage necessary to support the environment was reduced by 75%, aka, software-based architecture.

I applaud GENBAND for migrating to a soft UCaaS architecture without the end user realizing that GENBAND has migrated to an IP architecture. Mention is made that users are still supported in a full TDM environment. In my experience, gateways are still necessary to convert the IP backbone to TDM at the end points.
Steve Leaden 8/29/2017 4:13:52 PM

I pulled the following from an article on the GENBAND Website dated 12/6/12:

“There are about 40,000 legacy telephone switches in the world, and they still support about 900 million lines globally, about the same number of lines they supported just fifteen years ago. The phone services provided by these switches are of vital revenue and earnings importance to many of the world’s largest communications companies. These switches continue to be integral operating components in critical infrastructures across the globe, providing key communications services for consumers, businesses, first responders, health care, and military and governmental entities.

However, this critical infrastructure is on a rapid “countdown to shutdown.” Engineered for a 20-25 year design life, most switches have passed that mark and are now operating on borrowed time. Because many of these switches have performed in the past with reasonable regularity, operators can be easily lulled into believing that they will continue operating today as they did yesterday. This false sense of security is exacerbated by the attrition of skilled employees who operate these aging switches, the declining lack of vendor support, and a diminishing supply of spare replacement parts.”

I think we are saying the same thing on the most part, with your note that GENBAND continues to support customers with a full TDM architecture. The UC component applications are likely not integrated to the TDM end points, unless that is an exclusive offering from GENBAND.

And as I referred early on in the article, Centrex reliability has always been very high reliability level at 99.999%, and in my opinion, the gold standard for the industry.

It would be safe to say that not all carriers have supported a migration as GENBAND has.

I welcome your additional thoughts – feel free to email me at
Greg Zweig 8/30/2017 8:25:39 AM

Steve -- I'll connect off-line but I want to be clear that a carrier that had a DMS 100 can migrate that to a new GENBAND C20 (IP core) and maintain all the TDM Centrex assets the customer had on the DMS (the customer will likely never know the carrier's central office has been upgraded). Once that is completed the customer can selectively upgrade their Centrex to all or some IP services (mix and match). We absolutely allow the deployment of UC services with TDM endpoints. We make the infrastructure -- it is up to a given carrier to make a business decision to leverage those tools. Many do and continue to make upgrades. Your article suggests that since someone has an old phone on their desk they are inherently at risk -- that's not accurate at all. The customer's options are tied to the carrier they use and that carrier's business plan - not to the technology.

To Leave a Comment, Please Login or Register

UC Alerts
UC Blogs
UC ROI Tool RSS Feeds