Infinera would be better off allowing Transmode Systems to operate as an independent division than to integrate it much into its own corporation. Despite all of the assertions to the contrary, their cultures could not be more unalike – the former, a high-profile, Silicon Valley-type of company with people consistently logging long hours, oftentimes grappling with very quick deadlines – the latter, a Swedish, low-key, usually pragmatic kind of firm with historically fixed, lengthy timeframes for R&D completion, with a strong belief in the necessity for taking long vacations to recharge one’s batteries as well as to help maintain a relaxed demeanor in its development work. The biggest difference is that Transmode has been going after the sweet spot of the metro DWDM market – targeting the Tier II and below types of service providers and other customers in which there has not has been as much price competition because of less interest by the major optical vendors. (The size of these deals tend to be much smaller, and with a good number of newer networks unencumbered by the requirements of archaic architectures, Transmode has been able to maintain comparatively attractive gross margins.)
By leaving Transmode more or less alone, Infinera will avoid messing up a cash cow that could easily remain an entity unto itself principally on the European continent. It would help to gradually reduce Infinera’s debt load over time. Most vitally, the hope should be that with Transmode able to do its own thing, a good amount of the technical talent, who have been focussed on the metro from the bottom up right from the beginning, might remain with the vendor indefinitely.
Infinera should also not insist that Transmode get too far away from its comfort zone in backing some of its frequent rhetoric, such as the supposed advantages in moving to a large-scale, PIC-based technology or in placing so much of its attention on metro 100G, while hardly mentioning the still sizable 10G business. Regarding PICs, the proof is in the pudding in that no other systems integrator after all of these years has found it advisable to follow that product strategy. In addition, in initially deciding to move all the way up the food chain, the supplier succeeded in substantially reducing the size of the total DWDM space with its significant cost decline at a time when there was limited sales growth in the overall market.
As long as Transmode is more or less maintaining performance, there should be as little change as feasible in its cultural inclinations, which have turned out to be successful, including the deliberate way it develops its products. At least at one time, there was a rigid R&D strategy of introducing only two major product releases annually. It was a totally different mindset in that it was not necessarily about striving to solve a specific, present problem, but in trying to make sure that the targeted solution was done correctly during the course of the six-month program. (Several years ago, Transmode’s leapfrog to WDM-PON gear seemed to be an extreme example of this attitude.)
Whereas Transmode has had the luxury of tending to disengage relatively early from cutthroat pricing battles, Infinera has evidently been involved at times at by itself or along with other competitors to attempt to initially buy an account with the expectation that money would be made down the road with additions, etc. Perhaps Infinera has been too obsessed in its desire to show investors that its customers encompass x number of Tier I carriers as well as y number of mega-data centers. (One can only imagine what it cost the corporation to get Facebook to acknowledge publicly that it is a vendor to the social media business.)
Hopefully, the pickup of Transmode will be a step in a healthier direction for Infinera in that the discussion each quarter will move more towards the overall balance sheet rather than whether it has been able to penetrate a huge customer in some way — for it has been a matter in certain cases of big buyers keeping their incumbent suppliers honest rather than the firm really doing a substantial amount of business with Infinera. Also, while the Transmode product line will be able to become a part of an overall package with its long-haul solutions, it probably should never be at the expense of adversely affecting the cost structure of the metro player. In dealing with larger buyers of metro gear, which often demand everything under the sun, such as OTN switching functionality (Transmode has been more Ethernet-centric), it would probably be more advantageous for Infinera to look for other ways to provide these capabilities.
[written by Mark Lutkowitz]