For the second time in two weeks, Nortel Networks Corp. (NYSE/Toronto: NT) has issued a press release claiming the leading market position in sales of "optical Internet" gear.

But scratching the surface of both announcements reveals that Nortel's boasts, while valid, gloss over its continuing small share of the high-channel-count DWDM market. Further, analysts say Nortel must act vigorously to keep its overall place.

Let's take it from the top: In last week's announcement, dated February 19, Nortel says market research firm Ryan Hankin Kent Inc. (RHK) reported Nortel to have a 61.4 percent share of the North American long-haul DWDM market -- double the share that RHK gave it one year ago.

The overall North American market for DWDM in 2000 was $7.7 billion, according to RHK. The firm divides up the DWDM market into two segments, however, and Nortel doesn't dominate them both.

The first segment in RHK's taxonomy for 2000 devotes $6.1 billion of the overall DWDM market to systems with channel counts lower than 40 -- a segment that Nortel led with a 72 percent market share, according to RHK's research.

RHK's second segment includes so-called high-channel-count systems, with more than 40 channels -- a market RHK sizes to have been $1.6 billion in 2000. Nortel has just 10 percent of this chunk, which is led by Ciena Corp. (Nasdaq: CIEN) and NEC Corp. (Nasdaq: NIPNY).

Nortel's lack of presence in this division could be cause for concern, since that area is slated to be among the fastest growing in optical networking. For example, in publicity for a separate report on optical components, RHK cites "performance improvements -- such as higher channel counts" -- as drivers for triple-digit growth in DWDM systems over the next three years.

It's also worth noting that Nortel's share of the high-channel-count DWDM market didn't increase from 1999 to 2000 (see Nortel Spins Past Ciena).

The conclusion? Nortel must hustle if it means to keep its top spot in DWDM. It could lose out to other players quickly if high-channel-count systems start to predominate.

This analysis is borne out by research touted in this week's market share announcement from Nortel. On Tuesday, the company announced that research from The Dell'Oro Group shows Nortel to be top dog all 'round.

But report author Shin Umeda issues some caveats.

First, Dell'Oro does not use the same type of breakdown in its figures that RHK does. Instead, the firm sticks to sales counts it gets directly from the vendors in making its tallies. In the case of Nortel's long-haul DWDM figures, the vendor gloms together sales of its S/DMS Transport Node, an older product that works only with Nortel's Sonet add/drop multiplexers, and its newer Optera LH 1600, which isn't dependent on those ADMs.

It's an important distinction, some say. Nortel's products based on legacy Sonet and SDH (which Umeda estimates account for over half of all Nortel's optical revenues) are predicted to start phasing out as carriers demand next-generation gear with multivendor compatibility and higher channel counts.

Umeda says Nortel refuses to divulge which percentage of its long-haul DWDM revenues come from which products. Nortel confirms this. But a Nortel spokesperson says, "Optera LH definitely makes up the majority of our revenues in this space, and even without [the older products being included] we'd still have the number one spot."

For the record, Umeda says Dell'Oro has included only revenue from equipment in its reporting; no service and support is included. Also, he asserts that he's not double-counting DWDM equipment; what's counted in long haul does not appear in metro.

Other analysts confirm that Nortel needs to avoid complacency about its market position, in light of the changing nature of the DWDM market. "Nortel has a lot of old equipment that's intimately linked to its old-style ADMs," says Mark Storm, optical networking program leader at research firm Frost and Sullivan. "The [S/DMS] is where the bulk of the revenue is, and they're challenged to make the transition [to next-generation gear]."

He also says market researchers themselves need to revamp their taxonomies. "There's a key transition going on from Sonet/SDH to next-generation gear. It takes a while to slice and dice it all." He says researchers must avoid inflating the numbers by counting too much, while making sure not to miss the key trends by forcing new gear into old categories.

-- Mary Jander, senior editor, Light Reading http://www.lightreading.com