Column Column 5/25/2005
PALO ALTO, Calif. – Last week, we reported that Juniper Networks Inc. (Nasdaq: JNPR) was talking to Extreme Networks Inc. (Nasdaq: EXTR). Again (see Extreme Juniper Rumors Are Back).
Pretty much since Light Reading was born, we've done this story. Over (see Juniper's Slow Shopping Trip ). And over (see Juniper's Extreme Thoughts Are Back). Again (see Juniper To Buy Extreme? ).
But our reportage is not as silly as it seems. According to some sources here in the heart of Silicon Valley, the on-again/off-again relationship has pretty much worked as we described. Extreme CEO Gordon Stitt and Juniper CEO Scott Kriens periodically get together, sit down, have some coffee, and talk about the potential for a deal. After all, Kriens and Stitt are peers, former entrepreneurs working for companies funded by Kleiner Perkins Caufield & Byers. Both are IPO class of 1999. They know each other well.
In the early startup days, Kriens and Stitt were also both shepherded by ace Kleiner Perkins partner Vinod Khosla in the Japanese "Keiretsu" philosophy of creating an alliance of companies that could work together to build a virtual portfolio of best-of-breed point products to compete with Cisco Systems Inc.'s (Nasdaq: CSCO) offerings. (Interestingly, Kleiner Perkins appears to have removed its Keiretsu literature from its Website. Does it think the days of Keiretsu are over?)
The VCs have long since left Extreme and Juniper, cashing in their IPO gains, and Khosla's no longer on Juniper's board, but the concept of Keiretsu lives on. The long-term plan is finally panning out: Juniper is plunging heartily into Cisco's enterprise territory, and it now needs to acquire more point products to fill in the holes it has when matching up against Cisco's portfolio. It's already bought Netscreen, Peribit, and Redline. After years of tentative recovery, the smack-down match of the networking market has begun.
Richard Kramlich, general partner and founder of legendary Silicon Valley VC firm New Enterprise Associates (NEA), was an investor in Juniper. After all his investors' Juniper shares had been distributed, he stepped down from the company's board last fall. He says he is enjoying the Cisco/Juniper showdown.
"I always thought Juniper had the best product line in the space," says Kramlich. "But they were out-marketed and outsold by Cisco. The transition into enterprise and the acquisition of Netscreen is an interesting strategic move. Do I think it's tricky? Yes."
Most observers point out that Juniper has quickly assembled most of the key products it needs to go head-to-head with Cisco in the enteprise networking market. But it's still missing one key element: Juniper lacks a chassis-based Ethernet switch to compete with the Cisco Catalyst 6500, one of its best-selling products.
As Juniper hunts for Ethernet technology and a Catalyst 6500-killer, here are the acquisition options frequently mentioned: Extreme, Foundry Networks Inc. (Nasdaq: FDRY), and Force10 Networks Inc.
Foundry, led with an iron fist by CEO and founder Bobby Johnson, is the rebel child of the Ethernet switching market (see Bobby Johnson, Foundry Networks). It marches to the beat of its own drummer. And it's been profitable recently – Johnson holds a big stake and most folks don't think he will sell out.
Force10, say most reliable sources, wants a whole lot of dough. Kramlich, who also funded Force10 and sits on the company's board, maintains that the company is strong enough to hold out for an IPO.
"The team at Force10 is exceptional," he says. "They have the skill set to build a powerful array of products."
Juniper could build its own Ethernet switching product line, but that wouldn't come with a sales channel.
Then there's the Extreme option. As Extreme's share price erodes – victim of severe competition and pricing erosion in the Ethernet switching market – the deal becomes more likely. Juniper's lot is obviously far better than Extreme's, with it churning out healthy profits, well esconced as the number two player in what is essentially a two-man routing market, and sporting a robust valuation of nearly $14 billion to Extreme's $500 million or so.
Stitt has probably not sold because he's afraid of what this will look like – a fire sale. Such pride is understandable, but it's necessary to see the forest through the trees. The Ethernet switching market is not going to magically morph into some happy-go-lucky, high-margin business. Huawei Technologies Co. Ltd. and ZTE Corp. (Shenzhen: 000063; Hong Kong: 0763) are coming, and Cisco is turning the screws in the 10-Gig market. Sticking around by yourself in that space won't be much fun.
Kriens, of course, is worried that if he buys Extreme, his share price will get slaughtered – for the same reasons mentioned above. But this deal would be pitched as strategic. It's not about the margins; it's about giving Juniper a better enterprise sales channel and matching Cisco's products switch-for-switch. This is a case in which Wall Street might be too myopic.
The irony of it is that as Extreme's stock price falls, the deal becomes more likely – even if both Kriens and Stitt are afraid of how Wall Street will view the deal.
But is it really a bad deal? One Silicon Valley executive, whom I won't name, points out that with Extreme now trading at roughly 1.5 times revenue, that represents roughly the value of its revenue and sales channel, and that's about it. That's the valuation of, say, the dishwasher market. The products and technology come for free.
This also happens to be exactly what Juniper needs to build a solid enterprise sales channel: more customers, and an installed base of dishwashers – er, I mean, Ethernet switches.
Like forces building before the battle, Juniper's looking to put on more girth. It still needs more sales, more products, more revenue if it's going to seriously match Cisco in the enterprise market. For these reasons, I believe the Extreme/Juniper deal makes a whole lot of sense. It should be consummated this year. The Keiretsu classmates will finally be reunited.
— R. Scott Raynovich, US Editor, Light Reading