And then there were two: Cisco; Brocade and McDATA
The market can't sustain three players.
By Steve Duplessie, for Computerworld | Techworld | Published: 16:00, 17 October 2006
Q: I keep hearing the Brocade/McData deal is on and off (see "Brocade to buy McData for $713M"). As a McData user, I'm not sure what to think either way, but the Cisco rep must be smelling blood because they are all over us. What's your take? -- R.D., Research Triangle Park, N.C.
A: I love this question, and I have the answers -- for a change. The deal is not only good for everyone, with the possible exception of Cisco Systems Inc., it is necessary. Either it has to go through, or the U.S. government is going to have to bail out McData Corp. like it did Chrysler Corp., and I just don't see that happening.
First, understand the macroeconomics. The whole Fibre Channel business is about $1 billion annually. That's big if you compare it with my business, but insignificant when you view it inside the IT infrastructure market as a whole. That market is growing in little, teeny percentages. It will never, ever be a giant market. It will always be an economic niche, and as such, we have to face some realities.
First reality: At $1 billion, ultimately, there is not enough room for any more than two competitors. Look at the fiber switch market: Five years ago, there were 10 would-be players -- all about the same size -- most had gone public and had some cash, and all were ready for war. Over time, Darwinism did its thing, and the weak died or were absorbed by others because they had lost the path to market (storage resellers), and the market wasn't going to expand beyond that space. So there was no way to really compete.
That left us with two vendors from the old school -- McData and Brocade Communications Systems Inc. -- and one relatively new entrant that happens to be ginormous, namely Cisco. QLogic Corp.? Sure, it plays a small role in this sector, but it realized that if it wants to play in the switch market, it needs to create market opportunities. QLogic is the master of owning the cost play, but this market doesn't buy on just cost. Ninety-nine percent of the world doesn't use Fibre Channel -- and 90 percent doesn't even network storage -- so QLogic knows that if it can drag the costs down far enough, and make it easy enough, it can try to go after the broad market via value-added resellers and attempt to leverage its technology in ignored markets.
I like the play, but it will be a tough, expensive gamble, and the competition will be Ethernet. They aren't playing in the core market where Fibre Channel lives today.
McData and Brocade have dominated their respective positions in the market -- McData in the core large data centers with its directors, and Brocade is everywhere else with its switches. Both were delivered to the market as a component of a much larger solution by the storage resellers that won the business. Both were subjected to the fact that they were unable to get to market on their own, and as such, they were always at the mercy of the manufacturers that sold their stuff. Both have benefited and suffered when one major reseller decided to either give them an extra order at the end of a quarter or hold an order at the end of the quarter. Neither is big enough to really threaten its channel. The purchasing folks at the resellers and the big end users get paid to beat up vendors and lower costs. Even though the switch component of most deals was a very small percentage, if you incentivize someone to lower costs, they will try to do so wherever they can. Therefore, Brocade and McData ended up being forced t ccompete with each other and thus reduce the winners operating margin in order to satisfy short-term Wall Street expectations. It is just the way the game is played.
Unfortunately, what it means is that eventually one or both competitors will hit the wall and become financially unstable. Right now, that is McData, but a year ago, it was also Brocade. Meanwhile, our good friends at Cisco have all the time and money in the world to just wait it out.
Our government watchdog agency, the Federal Trade Commission, has nice lawyers and economists whose job it is to make sure that a deal like this is not going to be a net negative to consumers (i.e., cause harm by eliminating competition, reducing choice and potentially increasing costs). I have dealt with our friends on several such deals, and I must say how impressed I am with their desire to get to the bottom of things and to really do the right thing. I'm also impressed at whoever was able to educate them and get them to think in exactly the wrong way. I can speculate, but so can we all. I don't know what they will decide, but I do know that they are armed with much more data than they were, and that they appear to really want to do the right thing. The right thing is to let this deal happen, and here's why.
The market can't sustain three players. Cisco is not a variable; it's a given. It has the economic clout and power to wait however long it takes to be a player here. The perfect world for Cisco is that McData and Brocade die a natural death. Then no government or anyone else can think ill of it and will actually view it as a white knight. Cisco can then have the whole small pie to itself and reap the rewards that go with that without any entity yelling the bad words of monopoly or antitrust. At $2,000 per port, Cisco makes a ton of profit -- and good for it. Cisco could destroy McData and Brocade overnight if it so chose -- and if the government let it -- by simply giving away its Fibre Channel stuff for free.
In the fortune 1,000, for example, the biggest McData customer probably spends $1 million dollars in a great year on directors. The same customer spends $50 million to $100 million on Cisco IP gear. If Cisco gave the Fibre Channel stuff away, it would only represent a 1 to 2 percent discount, which would be meaningless to its bottom line. I know that, it knows that, and you now know that. It doesn't do that because it would be viewed badly by the feds, and Cisco is not stupid. It thinks, "Why lower prices to nothing when we can just wait, and then we'll end up taking that market naturally and be able to bask in the extra margin that the market affords us?"
What do the resellers think? The big guys should want the deal to happen. They know that they need a healthy two-player market, or else they are not only at risk if the one player has a yield problem (imagine not being able to ship and recognize your million-dollar deal because a $1,000 part didn't show up), but also IT always wants a second source. Resellers don't care about the fact that they may lose some leverage on pricing because it won't affect the overall deal. They do care about the fact that all the resellers are bigger and much more powerful than Brocade or McData, so even if the entity combines, they can still dictate most of the rules.
They can't say that about Cisco. They know they have to play with the "big C," but they also aren't dumb and realize that Cisco has much more clout than they do in most accounts. Even the big systems vendors have to be weary of someone as powerful as Cisco because when it makes a market move, it could dramatically affect your world. If you are IBM or EMC Corp., you have placed a lot of McData directors in the world where customers are going to look to you to fix the problem, and that will not only mean they will have to bring Cisco in to replace McData but it will also cost them a lot of money and result in a loss in leverage. Cisco already has a big purchasing contract with the customer.
What should users think? Users need to accept reality. If these two don't get together, then it is only a matter of time before you will be forced to make a move on one of them anyway. If they do get together, they at least have a fighting chance, and for the short term, they look good financially. As long as Brocade does not force McData users to change systems -- and I have no reason to believe they would, because it would be suicide -- and can now support the platform with much more financial resources, users should be happy. McData is installed in really big shops that really don't want to screw with something that critical, especially when it works. Cisco is already taking share over the past few quarters, and the company seems to have figured out the recipe. It isn't going to stop now. It's going to get more aggressive, and why shouldn't it?
So I am not saying that this deal will always guarantee users choice. I can't say that a combined company won't ever screw up. I can't say that even if a merger does happen, users won't use this event as a reason to leave their systems and move to Cisco. I can say that this market won't support three suppliers. I can say that without a bailout, I can't see how McData can survive. I can say that even if someone else bought McData, by the time the company was fixed, the insecurity the customer base feels will have enabled Cisco to take enough market share away that there will be no way to recover. Customers in this market are risk-averse, and no one thinks Cisco is going out of business. It would rather give up all the benefits a free market with healthy competitors gives it in exchange for knowing that it has eliminated risk at this layer.
I also know that if the deal doesn't happen, we'll be left with Cisco and Brocade, and Brocade will be able have a healthy business that will slowly decline but can be managed to be profitable for a long time until Internet SCSI (Ethernet) takes away the departmental, smaller Fibre Channel server market altogether -- and guess who wins in that battle?
Managing costs is not the same as investing in research and development to better compete in a growing market. Users benefit by the investment their vendors make in R&D to build them better products. R&D costs money. If I'm only managing costs, I can't invest in new development. If I'm the only person left in the market, I can invest in R&D, but boy, are you going to pay for it. EMC Corp. has sold billions of dollars worth of Symmetrix Remote Data Facility because once it was successful, its investment in R&D resulted in a product the market needed and couldn't get anywhere else for many, many years. I don't blame EMC; I applaud it. There are examples of this every time someone dominates a market.
So ultimately, the market as we know it will cease to exist for Brocade. That might be in 10 years or more, but someday, it won't make sense to have a specialized network that costs more once we can solve those problems with a standard network that costs much less. So Brocade will have either moved into other higher-value businesses, which it's already doing, or the company will go away. Brocade knows that it can't compete on cost in the long term, nor does it want to be in that position. The company wants to leverage its success and footprint by adding value that it can monetize; Brocade isn't betting on plumbing.
McData and Brocade would have a defensible customer base where the united company would be the dominant player -- at least for now. If they can defend that base and bring value on top of it, they have a great chance. The most expensive and hardest aspect of any company is customer acquisition. Once you have the customer, you have the upper hand. If you can build things they need and want, you can make money. If you don't, you die. Sooner or later, no matter what, the plumbing will become more and more standard and whoever who can build it cheapest will win.
I'd love to hear your thoughts on where my logic fails. I love Cisco, but the world will be a better place if there is an alternative. I can't think of one scenario where this deal isn't better for all, except for Cisco -- but it can wait.