How common is it for companies to ship products to customers


those products have been formally introduced?

Probably all


common, but that question is central to the rapidly evolving (and increasingly tense) story surrounding

Ancor Communications


, which is trying to snare share from industry leader (and


Red Hot)

Brocade Communications


in the super-hot fibre-channel switch industry.

As this column has

previously reported, Ancor recently won a coup when it signed

Sun Microsystems

(SUNW) - Get Report

as a customer (in return for granting Sun warrants on its stock). What hasn't been generally reported, however, is what happened last week when Ancor reported earnings. Deep in its earnings release, Ancor matter-of-factly said that it had started shipping a "next generation" eight-port switch last quarter. Don't get caught up in what an eight-port switch is -- that's not important. What's important is that the company later said that the new switch comprised everything shipped to

MTI Technology



Herb's Latest: Join the discussion on


Message Boards.

This is where things get dicey. Shipments of this new switch, according to analysts, accounted for 37% of Ancor's sales in a quarter that handily beat analyst expectations.

What's more, while sales from the prior quarter rose 11%, receivables soared by 70%; results like those often mean customers have been given incentives to take more product than they otherwise might accept. The hint by a number of short-sellers is that by shipping the new product, Ancor helped prop up what would've otherwise been a lackluster quarter.

Ancor, however, says that's not the case. CFO Steve Snyder says the higher receivables reflect the timing of MTI's order, which was late in the quarter. And as for shipping product that hasn't yet been announced, he says: "It's very common for OEM suppliers to do that -- or even those who ship to end users. It is quite a common occurrence to put product out in the field before it is publicly announced. This public announcement is really timed around other events. In this case, we are timing it around an industry event that takes place in early November, so there's nothing really unusual there."

Not so (surprise, surprise), according to Brocade CEO Greg Reyes. (Like you expect him to agree with Ancor?) "It's highly unorthodox to recognize revenue for new products that have not been publicly introduced or proven themselves to be viable new products," he says. He adds that Brocade goes one step further and doesn't recognize revenue until OEMs ship products out to end customers. For example, he says, Brocade shipped a new product to



last April, but didn't recognize that revenue until July, when Compaq starting shipping the product containing the Brocade switches.

Why does any of this matter? 'Cause yesterday



, a competitor of Ancor and Brocade, skidded 20% after reporting its first quarterly results as a public company. Hard to say where the disappointment was, but interestingly, Vixel has a market cap of $635 million and last quarter had revenue of $8.9 million (and it's getting clobbered!). That compares with Ancor, which last quarter had a market cap of $925 million and revenue of $4 million. (Both are puny when compared with Brocade, which has a market cap of $7.3 billion and whose sales last quarter topped $20 million.)

Oh, and one other thing: Brocade hired


, a respected testing company, to test its switches against Ancor's and Vixel's. The results were overwhelmingly in favor of Brocade. Ancor President Cal Nelson dismissed the test as an unfair comparison, because it tests Ancor's older switch, which isn't the one going to Sun or MTI. "It is totally ridiculous," he told my associate,

Mark Martinez

. "It isn't even the product that they are competing against. In our view, that is a misrepresentation. It is clear that Brocade is trying to manipulate the marketplace. They know that this is not the product that we kicked their butts with. They know this."

Reyes' response to me: "Great, then let's put the allegedly new switch through KeyLabs' test. For months they said theirs was better than ours. They should've qualified their statements because they've been making misrepresentations for months."

Adds KeyLabs CEO J.D. Brisk: "We run a lot of these types of comparisons, and most of these are not published. And Brocade can have this information released or not, depending on how it comes out."

Brisk adds that companies come to KeyLabs with the hopes of looking good. In the end, "they pay us to tell them whether or not the baby is ugly. Sometimes it looks good, and sometimes it looks like a chimp. We don't care one way or the other -- it is simply data to us."

If a company doesn't like what it hears, he adds, there's little chance that anyone will ever know because the results will never be publicized by the company that requested the test in the first place. "We executed the task that we were paid to do," Brisk says. "What a company does with the results is their business. But we are going to call them as we see them."

Amazonian Ambush

Amazon: Join the discussion on


Message Boards.

From one of this column's

(AMZN) - Get Report

watchers, after listening to Wednesday's conference call: "I thought the biggest surprise was the comment about portal relationships. I think it was

CEO Jeff Bezos who said, almost at the end of the call, that Amazon may not renew some portal relationships when they expire. If mighty Amazon is questioning the value of those relationships, what does that say for the future ad streams of




America Online


, etc.?"

Good question!

Herb Greenberg writes daily for In keeping with TSC's editorial policy, he doesn't own or short individual stocks, though he owns stock in He also doesn't invest in hedge funds or other private investment partnerships. He welcomes your feedback at Greenberg also writes a monthly column for Fortune.

Mark Martinez assisted with the reporting of this column.