Kevin, thanks for your kind words.
The comments below were written almost a year ago. Within the context of a bullish stance, I highlighted concerns about Infinera’s competitive advantage in the Metro Aggregation market. My thesis boiled down to: If (1) the photonic integrated chip (PIC) is Infinera’s “special sauce”, and (2) PICs are best-suited to high-throughput environments, and (3) metro aggregation - especially at the edge - is not high-throughput, then how will Infinera integrate Transmode such that it exhibits competitive advantage equal to Infinera’s?
While I think we haven’t seen the answer fully play out yet (the acquisition is too recent for design efforts to have come to fruition in production models), I’ve also reached alternate conclusions…
First, recall that - prior to the acquisition - Transmode’s margins were higher then Infinera’s. Transmode obviously has some “special sauce” of their own. Second, while the PIC is critically important, we shouldn’t focus on it to the exclusion of all else. Infinera management has shown extraordinary vision. They are “skating to where the puck is going to be”. I think their penetration with the ICPs (Internet Content Providers - today’s customers) rather than the Tier 1 telcos (yesterday’s customers, although still seen as critically important in some circles) speaks to this, as does the amount of time that Cloud Xpress has been on the market without credible competition. Frankly, I’m starting to put “vision” on almost equal footing with “PIC” in the moat department… All that said, here are my thoughts from a year ago.
Concerns Regarding Infinera’s Competitive Advantage in the Metro Aggregation Market
My first concern is with the metro aggregator customer base. Who are the customers most likely to buy Infinera’s (and/or Transmode’s) equipment? In the long haul market, telecom providers have long been a dominant factor, but a set of newcomers has eroded their former dominance. These include cable companies, Internet content providers (ICPs), cloud services providers, and the “wholesale and enterprise carriers” (formerly known as Prince – oops! I mean, formerly known as “bandwidth wholesalers”). Will the metro aggregation market feature the same group of industry players? We probably won’t know this market’s customer profile until we start to see orders roughly a year from now. Do any industry insiders reading this message board have a sense of how the list of players in metro aggregation might differ from long haul? In the meantime, I have at least one data point for an educated guess. Included in the slide deck that Infinera used to present the rationale for the Transmode acquisition was a customer comparison. The Transmode list seemed heavily weighted towards telecom providers, as best as I could tell (CFO Feller did mention that Transmode had strong business with cable operators too). I raise this issue because domestic Tier 1 telecom providers have not, shall we say, been instrumental in Infinera’s long haul success. If they form a critical buying group in the metro aggregation market, Infinera may find it a tougher market to penetrate. A second data point is that Infinera touted bi-directional cross-selling opportunities as a benefit of the Transmode acquisition. Those benefits would be muted if there is little overlap between potential long-haul and metro aggregation customer lists.
My second concern is that I see Infinera’s PIC as one of its greatest strengths. But it isn’t clear to me that PIC technology will have as big an impact in the metro aggregation market as it does in the long-haul market. I base this concern on a point raised by co-Founder Dr. Welch, but before I talk about what he said, I think it may be useful to speak to some of the slides from the Transmode acquisition conference call, and how those slides partition the metro aggregation market. I describe how to access links to these slides in the “Resources” section several paragraphs below this one. The slides present Metro Access, Metro Edge, and Metro Core as the submarkets that make up Metro Aggregation market. While some larger, data-heavy enterprises might feed network traffic directly into the Metro Edge, typical business and consumer data feeds into Metro Access, where their network traffic is aggregated and then fed into Metro Edge. I think that further aggregation occurs at the Metro Edge to feed the Metro Core (I could be wrong – management didn’t speak to these slides in detail). The Metro Core, in turn, interfaces with the long-haul part of the network. Dr. Welch’s comment was “Photonic integration has been exercised to this point in applications that require high bandwidth and high capacity. … When you get out to the client access side, there is less of an opportunity to use the photonic integration technology. When you get up to the aggregated ring, sort of capacities increase, and there is a greater opportunity to do so.” (from the Transmode acquisition conference call) If I understand this comment correctly, PIC technology would be mostly inappropriate for Metro Access and might be appropriate for Metro Edge (and seems very appropriate for Metro Core). Based on this, I think we need to watch to see if Infinera’s anticipated metro aggregation product is PIC-based. If it is, we’ll want to understand what submarkets within the metro aggregation market it might serve. When management offers us size projections of the total metro aggregation market, it would be good to understand the size of the subset this product will serve. If the anticipated product is not PIC-based, then what makes it excellent?
I want to be clear about my use of the word “concern”. These are areas where I’ll want to focus my attention as we all learn more about the metro aggregation market. I hope that helps. If you have concerns of your own, please post them so we develop a solid list of issues to monitor as management goes into more detail about metro aggregation over time.
Thanks and best wishes,
TMFDatabaseBob (long: INFN)
Peace on Earth