Why Did Juniper Really Ax the DX Line?

The company says it's about commoditization, but others says the problem was Juniper's missteps.

A series of missteps and a rapidly shifting market-and not a space that's moving toward commoditization-may lie at the heart of Juniper Networks' decision to pull the plug on its DX line of application delivery controllers.

When Juniper officials in late January announced they would discontinue its DX line, CEO Scott Kriens told eWEEK that the market in which the products played was commoditizing, and that Juniper's strategy is to pursue more high-performance opportunities that can command more premium pricing.

However, industry observers disagreed, saying Juniper failed to execute on its acquisition of Redline Networks, did not fully exploit the competitive advantages in the acquired DX product line and saw F5 Networks unexpectedly rise quickly to the dominant position in the market.

Juniper's exit, three years and at least $132 million later, raised more questions than it answered. Why didn't the DX line keep pace with market growth rates? What happened to the planned integration of the DX line with the WX line of WAN optimization controllers? Also, will not having a viable product in that market hurt Juniper's bid to gain a greater share of the enterprise market?

"The whole statement that [the market] is going commodity is just absurd. We're seeing more innovation and value," said Gartner analyst Joe Skorupa, summing up the reaction of several sources. "Juniper failed to execute in nearly every aspect of the DX business."

"We're not seeing commoditization," said Jason Needham, senior director of product management at F5 Networks. "The traditional application delivery market is really strong."

Not to mention competitive, thanks in part to a big bet F5 Networks made several years ago that is providing the company with outsized returns. F5 had embarked on a re-architecting of its operating system when it moved from Layer 4 load balancing to Layer 7 load balancing, and made its operating system more modular, allowing it to add new features at a much faster pace.

With the launch of its TMOS 9, and an innovative developer community program built around its iRules technology, F5 was able to catch up with the benefits unique to the DX product line.

"F5 and others had developer [momentum]. Redline had happy customers and smart guys working on the product. But it started to get marginalized as a niche opportunity," said one source close to Juniper, who asked not to be identified.

That was not the case when Juniper bought Redline in 2005. Then its application delivery controllers were considered best of breed in a market that enjoyed robust growth rates.

However, Juniper's track record at integrating acquisitions is a spotty one, and the Redline acquisition proved to be one of the harder ones for Juniper to assimilate, said the source.

"Juniper hasn't done a good job at getting broader leverage from their acquisitions. They aren't good at taking a good point product and integrating it into a broader portfolio for more competitive advantage. They weren't able to do that with Redline," said the source.

Others said that while Juniper is highly skilled at selling superior networking technology to service providers, its transition into the enterprise has been halting at best.

"I think that all of the instincts from a sales and marketing standpoint that made Juniper very successful in the service provider space didn't do them any favors when it came to enterprise marketing. They made very smart technology acquisitions, and yet it just seemed problematic to keep the marketers that came with those technologies engaged. Enterprise and service provider [markets] are very different and counter-intuitive. The consummate sales and marketing [personnel] for service providers were in learning mode when it came to the enterprise," said another source close to Juniper, who also asked to remain anonymous.

At the end of the day, Juniper realized that to stay competitive in the market would have required a greater investment than the company was willing to make.

"I think it was a financial decision. The technology was solid, but they didn't really differentiate themselves from competitors in a substantial way. They would have had to really invest to get in the upper right hand quadrant [of the Gartner Magic Quadrant]," said Juniper channel partner Dave Gilden, a partner at Acuity Solutions.

"They faced a cross roads with the DX because of the way the market evolved around them. They no longer had a best-of-breed product. Being the number four or five player doesn't fit well in Juniper's strategy. It would have required a big investment. My guess is they decided to focus on what's critical to the business," said the first source.

But will Juniper's lack of presence in that market hurt its broader bid to become a significant player in the enterprise market? At least one observer thinks so.

"Not having an application delivery controller leaves a real hole [in Juniper's product portfolio] because those are becoming more strategic," said Gartner's Skorupa.