In the past, we have provided countless examples of the dominant ICPs demonstrating an unsatisfactory amount of concern for the long-term viability of the optical ecosystem. One of the most damaging practices has been the emphasis on OLS, which allow these customers to selectively pick and choose various piece-parts of different system vendors in order to achieve the lowest cost per bit possible in the aggregate, including less expense somewhat operationally, such as with the least amount of power usage. In our opinion, it is quite clear that ADVA Optical Networking felt forced to buy MRV, despite at a minimum, the short-term headaches of having to support a second platform (which has hardly been a revenue/profit powerhouse) as well as the burden of taking on debt, so that it could obtain unique transponder as well as optical plug solutions for its current FSP 3000 CloudConnect offering. We believe that there is a tremendous amount of pressure on ADVA to gain a "time to market" edge over "a number of competitors that have similar solutions" in order to increase its share of the data center interconnect space, in the face of even less impressive margin generation in not necessarily offering a complete system to such buyers.
Before the large data center companies represented a large portion of the optical space, ADVA would have never considered an OEM arrangement with another metro player, let alone a purchase. It is very proud of its engineering prowess (with a great deal of justification). However, more recently, these relatively new customers have grabbed some of its talent, which makes it even more difficult to meet desirable dates for functionality.
So, we have not found ADVA’s stress on other reasons for the MRV purchase to be compelling including its contribution to vendor consolidation. While ADVA certainly has MRV’s customers to theoretically penetrate with its other products, it is hard to imagine that the “common, mutual architectures” discussed during ADVA’s last quarterly earnings call will ultimately have much to do with the OptiDriver, which could only serve to potentially alienate those buyers. In addition, while there would certainly be some “leveraging [of] their engineering teams, “their sales organization” would be a whole other matter.
ADVA’s leadership was initially wary about getting too far into the hyperscale vortex. On the same conference call, the CEO stated: “[A] year and a half or two years ago, I said I don’t even want more than 10% of our business to go to ICPs, because of how good they are in understanding, constructing…and designing their own products and all those other pieces, so it was a balance. And then our business grew a lot faster…, which was definitely good for us, but the fact is when you look at the business as a whole, it is strategically important for us, [but] margins aren’t going to be high ever in those areas.”
So, for now, ADVA is stuck in a portion of the market, which will come up with even more imaginative and unprecedented ways of squeezing ever more costs out of both optical system and component suppliers, to perhaps the point in which it could conceivably do damage one day to getting the necessary technology developments for their own networks.
[written by Mark Lutkowitz]