Juniper feels growing pains

Cisco rival struggles with enterprise, slowing growth in core routing

Juniper Networks continues to challenge Cisco but faces growing pains in the enterprise network and core routing markets.

Juniper Networks is at a crossroads.

The 11-year-old company, formed to challenge Cisco in service provider core routing, is facing slowing growth in that market (The company is announcing a new core router Monday that it hopes will boost sales.)


Slideshow: Juniper Networks acquisitions fuel its enterprise business


Its entry three years ago into the enterprise market – also a Cisco stronghold -- has yet to result in profits after an investment of at least $5 billion acquiring several smaller companies. Continued sales and marketing and R&D investment in enterprise network technology is dragging down Juniper’s overall profits, analysts note.

Fortunes are brighter in service provider edge routing, where Juniper is regaining some market share after five consecutive quarters of losses. The company’s share in edge routing fell from 24% in the third quarter of 2005 to 14% in the fourth quarter of 2006, but has climbed back to 18% in the first quarter of this year, according to Dell’Oro Group.

A stock option backdating situation and the recent exodus of several top executives also downcast the company’s image, if not its momentum. Juniper is still without a CFO after losing Robert Dykes to an ad monetizing start-up in mid-March.

The company has completed options backdating investigations – it was one of 80 to 100 companies ensnared in that dragnet – but it cost Juniper $900 million in charges and the restatement of several quarters of earnings.

And just last week, UBS Warburg downgraded Juniper stock based on its ballooning valuation.

Overall, the company is growing – at $2.7 billion, 2007 revenue expectations are 17% better than last year’s; but clearly, Juniper is experiencing growing pains, too.

“The company has grown up very rapidly in terms of revenue, but in a number of other areas we’re not as effective as we could be or want to be at this scale,” says Juniper COO Stephen Elop, who took on the newly created position five months ago. Elop came to Juniper from software developer Adobe Systems, where he was president of worldwide field operations.

“Juniper’s challenge (is) too many rabbits to chase with too few resources available,” states CIBC World Markets analyst Ittai Kidron in a recent bulletin on the company’s first quarter.

One of Juniper’s challenges is in building awareness of the company’s strengths and quality of products, Elop says. Another is the ease with which companies can do business with Juniper – he says Juniper could improve its customer relations and transactions.

“It’s that class of things, that need to focus on execution that is very much at the heart of the opportunity that exists to take Juniper to that next level,” Elop says. “That’s why I was hired into the company, why this role was re-established.”

Elop says he was attracted to Juniper for its innovation, competitive passion and fire, and its ability to gain an “unfair share” over competitors of the markets in which it competes. Right now, however, the enterprise market may be unfair to Juniper.

Juniper’s enterprise angle

Juniper’s initial thrust into the enterprise market was the $4.2 billion acquisition of VPN and firewall stalwart NetScreen Technologies in 2004. Juniper not only acquired a leading portfolio of enterprise security products but an installed base of customers and channel resellers with which to grow from.

Juniper soon followed that acquisition with just less than $1 billion in others in 2005 to fill out its enterprise portfolio: WAN optimization vendor Peribit Networks, for $337 million; application acceleration vendor Redline Networks, for $132 million; and network access security vendor Funk Software for $100 million (see slideshow).

Technology from these companies makes up Juniper’s Service Layer Technology (SLT) group, one of two main product development and marketing groups within the company. The other is the Infrastructure Products Group (IPG), comprised predominantly of service provider routers and switches.

But Juniper has yet to see a profit from its three-year, $5 billion investment in SLT. Much of that is because of the company’s continued investment in building around those acquisitions with sales and marketing expenses – including ongoing channel development – and R&D costs associated with complementary enterprise product development, such as the J-series and Secure Services Gateway branch office routers. Analysts and other industry watchers also expect Juniper to unveil LAN switches later this year.

And even though SLT’s revenue is growing in the mid-teens, it’s not keeping up with the 20%-plus operating margin and 35%-plus revenue growth that NetScreen was generating before Juniper acquired it, according to UBS Warburg Analyst Nikos Theodosopoulos. If SLT does not turn a profit this year, Juniper should reconsider its overall enterprise strategy, he says.

“We do not believe Juniper’s Enterprise business can achieve consistent double digit operating margins and double digit revenue growth,” Theodosopoulos stated in a report on Juniper’s first quarter. “Given… the backdrop of increasing shareholder activism, private equity and [merger and acquisition], there is likely to be increased pressure on Juniper management to show improved results or new strategic direction in the enterprise market by the end of 2007.”

In a bulletin announcing UBS’s downgrade of Juniper’s stock last week, Theodosopoulos states that Juniper management expects SLT to turn a profit this year. But that will be through an “intense” focus on expense control and sales force execution, he believes -- “the results of which are yet to surface.”

Elop would not confirm that timeframe but said the unit should turn a profit in the “near-term,” which he defined as “quarters, not years."

“Things are going very well there and heading in the right direction,” he says. “Quite frankly, we’ve got to win some market share, make sure we’re focused on the right elements of the enterprise opportunities, tighten our focus in some cases…we’re very confident in terms of the opportunity that’s presented there.”

Some analysts still need convincing.

"Their strategy had been to buy best-of-breed technologies at various points in the network,” says Zeus Kerravala of the Yankee Group. “They built a router; they bought a security product; they bought a WAN optimizer and they bought a load balancer. What is it about those things that go together? There’s no real story you can weave together. Until there is, I think they’ll always struggle in the enterprise.”

Elop says there have been no discussions internally on divesting SLT and the enterprise operations in whole or part should profitability escape Juniper’s “near-term” window. He says Juniper will continue to focus on the “high-performance networking” requirements of large enterprises, financial institutions and the public sector.

“We are committed to the enterprise, it’s a critical part of our growth strategy so there’s no need for strategic considerations,” Elop says.

Core routing plan

Core routing also remains a critical part of Juniper’s growth strategy but the company’s share in that market declined from 37% to 30% in the past year, according to Dell’Oro Group. The research outfit cited no new system shipments or customers – 2,500 T640s have been deployed in production networks since the product’s debut in 2002, according to Juniper -- perhaps because they were awaiting the T1600, Juniper’s next-generation core router unveiled Monday.

The T1600 is intended to scale customers of Juniper’s previous high-end, the 5-year-old T640, further into the multiterabit range. Juniper has high hopes that the T1600, which is scheduled to ship in the fourth quarter, will eventually end its market share slide.

“When we designed the T640 and the TX [switching matrix for interconnecting T640s] we designed into those products scalability that we are now going to start to deliver on, which again will give Juniper technology leadership,” says Kim Perdikou, executive vice president of IPG and general manager of Juniper’s service provider business team. “Watch this space and track what the customers do.”

One potential customer, however, already bypassed Juniper as a replacement vendor for its core. AT&T selected Cisco and its CRS-1 platform as a substitute for Avici Systems after Avici announced it was exiting the core router market.

Perdikou was philosophical about the snub.

“I have a competitive streak that I didn’t realize how strong it was,” she says. “Any deals that we’ve focused on that we don’t win I certainly want to learn from. It’s the networks that get built, how they get built and how many products that the customer buys that’s really important.”

As for core router growth overall, analysts say the market will be flattening for everyone for awhile because of four years of double digit annual growth rates, sales of less expensive Ethernet ports, and sales of high-speed ports outpacing full system revenue.

Juniper on the edge

In edge routing, two of Juniper’s three key platforms – the M120 and E320 – contributed significant revenue in the first quarter. Much of the E320’s, though, was deferred revenue from Verizon.

Juniper also realized initial sales of its new MX960 Carrier Ethernet switch in the first quarter.

“I am much more positive” in edge routing, Perdikou says. “We’re going in the right direction.”

“The M120 is doing very well, it’s the right size and the right product for where it’s going,” says Eve Griliches, a telecom equipment analyst at IDC. “They’ve got a lot of customers who’ve waited for the MX960 as well.”

They’ve waited a while. One of the knocks on Juniper over the past two years was its dismissal of Ethernet switching as a key architectural element in emerging service networks, such as IPTV, and its lack of product breadth and integration expertise to pull all of those elements together.

Juniper suffered for it. Fairly or not, it lost share to Alcatel in IP edge aggregation routing – though some claim Alcatel’s inclusion of its 7450 Ethernet Services Switch in those numbers was misleading. Juniper was also losing the PR battle as analysts and media continually made note of Alcatel’s specific gains in IPTV because of the inclusion of an Ethernet switch, integration know-how, and presence in high profile projects such as AT&T’s Project Lightspeed.

Perdikou says the IPTV backlash was caused by a disconnect between what Juniper customers wanted and what analysts thought the company should sell them.

“Our customers do not require Juniper to sell them an end-to-end solution for IPTV,” she says. “What the customer is really looking for is a best-of-breed product that fits into an IPTV solution and that the vendor understands how it integrates with other products in the network. There was a very strong feeling – correctly so -- we were outmarketed and someone else had set the agenda for the IPTV market. We were in IPTV networks that were working; we hadn’t marketed as well as we could.”

So marketing remains one of Juniper’s challenges, in addition to:

•  Squeezing profits from enterprise.

•  Winning back share in core routing.

•  Keeping the momentum going in edge routing. . .

…all while scaling the company to grow to beyond $3 billion next year.

“Technologically, they’ve had a very strong advantage for a long time,” Griliches says. “I don’t think the liability of being a product line-centric company is a detriment to them at all. Ultimately, it ends up being an asset.”

Learn more about this topic

Slideshow: Juniper Networks acquisitions fuel its enterprise business

06/11/07

Juniper unveils giant router

06/11/07

Slideshow: Juniper unveils the T1600

06/11/07

South Carolina probation department runs on Juniper enterprise gear06/14/07
Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.
Now read: Getting grounded in IoT