Claiming the leadership position in any market is always a feather in a company's cap, but if the product market under discussion is PBX systems (aka IP telephony systems), declaring a market leader is somewhat problematic. You may have read my article, "Mid Year Voice System Market Update: Cisco Still #1," in which I designated Cisco Systems the current North American PBX market leader based on line station shipments. But based on recent analysis I may have been wrong. Then again I may also have been right, depending on how you define PBX line station shipments.Why, if most of the leading system suppliers distribute quarterly shipment data updates to industry analysts, is there now a new problem determining market share estimates?
Each company reports its data differently from the others. A few years ago, most of the companies adhered to a similar reporting policy: PBX shipment data at the system and line station levels segmented by cell sizes (i.e., line size segments, e.g. 100-400 lines, 400-1,000 lines); the same for Key/Hybrids. As an added bonus, the line station shipment data was usually segmented as either new or add-on. Geographic breakouts have also been reported, although some companies reported North America data, embedding their U.S. market shipments into that figure.
Today, most companies no longer differentiate between a PBX and Key/Hybrid if it is an IP telephony design, and cell size segmentation has mostly been abandoned. A few still report shipment data for new and add-on line stations, but a growing number don't. Almost all continue to provide geographic breakouts. It should be noted that Siemens Enterprise Communications Group, currently a second tier player in North America (despite being a first tier global supplier), has not reported shipment data for many years; one hopes that the new Gores Group owners will address this situation and provide some reporting data. Inter-Tel, before its acquisition by Mitel, also failed to report, but its numbers are now included in Mitel's report.
Not reporting data or reporting indecipherable data becomes a major problem, however, if the company is a first tier supplier with a significant impact on reported market conditions. Beginning with its just released Q3 2008 shipment report, Nortel has thrown a new wrinkle into the mix. For the past several years, Nortel reported new system shipments only, and declined to report line station shipments. To its credit, Nortel segmented the system shipments by product (Norstar, BCM, Meridian 1, CS 1000E, CS 2100). Line station shipments had to be estimated by product category, which reduced the accuracy of the overall market data.
The latest quarterly report no longer includes system shipments nor product segmentation nor geographic breakouts. Instead, Nortel reported but a single number, representing total line station shipments for all products on global basis.
I have no major issue with Nortel bundling its PBX and Key/Hybrid data, because other competitors are also guilty of this to some degree, and Nortel has indicated it may yet segment the data by geography, but a larger problem remains. The company is on target to ship about 10 million line stations this year based on how it counts line station shipments-a figure that dwarfs the competition. For example, Cisco and Avaya are both tracking for approximately 6 million line stations. So if Nortel's data is taken at face value, the company would find itself at the top of the market leader board.
Based on a personal conversation with Nortel representatives, the single reported figure includes installed Meridian 1 PBX line stations migrated to Nortel's current generation CS 1000E IP telephony system model. Less significantly, it also includes line stations migrated from its old TDM-based SL-100 model to the IP-based CS 2100 model. The reported figure includes all existing analog, digital, and IP stations migrated as is from the old to new design platforms.
For an installed Meridian 1 system, an upgrade to a CS 1000E requires the following, at minimum: replacement of common control and common equipment hardware; replacement of digital T1 trunk circuit cards; a generic software upgrade; and installation of new voice terminals (desktop instruments, soft phones) as specified. Most Meridian 1 port circuit cards and installed telephone instruments can be re-used if the enterprise opts not to forklift-upgrade the entire system.
Nortel believes that counting the line stations migrated between system platforms is an accurate and fair representation of current day shipment activity, because a customer has a choice whether or not to upgrade an aging Meridian 1 system to the current CS 1000E model, or potentially replace it with a competitor's solution (in which case all installed line stations would be included in the reporting data for the competitor that won the business).
The problem with Nortel's approach is that other system suppliers, such as Avaya and Mitel, are in a similar situation, but do not report migrated line stations. They do report add-ons or changed line stations, e.g. replacing a digital telephone with an IP telephone. Existing analog, digital, and IP lines that are migrated forward are not included in the reporting data by these vendors.
Avaya customers can migrate Definity common equipment and telephones dating from the 1990s to the current S8500 or S8700 models; Mitel customers can likewise migrate SX-2000 Light PE common equipment and telephones to the current 3300 IP model. For an extreme example of system migration, Ericsson MD 110 PIM equipment installed more than two decades ago can be migrated to and integrated with the supplier's current generation MX-One model.
Now comes the conundrum: How does a market research analyst reconcile Nortel's new reporting data with the reporting data from other system suppliers? Also, how are we to determine geographic breakouts for the data? Since Nortel's previous line station data were estimates, should we create new estimates from old estimates?
Why are Nortel's shipment reports going from bad to worse? Does Nortel believe market research analysts will now crown the company the market leader without sufficient segmentation detail? Is it just a coincidence that Nortel has decided to change its reporting methodology at a time the company is facing severe financial difficulties? If you haven't noticed, Nortel's stock share price during the past few years has tumbled from an adjusted price of more than $1,100 (taking into account a 1:10 reverse split) to less than 50 cents at close of business on November 21, 2008.
Can Nortel actually be the PBX and total CPE market leader based on line station shipments? Are they actually doing better than Cisco Systems, even as Nortel's Enterprise Solutions business unit is reporting negative operating margins?
As I will soon be drafting my annual market review article and preparing my slide materials for VoiceCon Orlando 2009, I will be forced to make a decision on how best to report Nortel market activity compared to the competition. It is not a task I look forward to, because regardless of how I report Nortel data (at face value or self-created estimates) someone is not going to be happy, unless Nortel gets with the program and provides a greater degree of granularity with its reporting data.