Optimizing the WAN Optimization Space

by: Anil Sharma

When IT applications from a company’s headquarters are rolled out to the branch offices to provide seamless work environment across the whole enterprise, data transfer needs on wide area network links between branch offices and headquarters increase manifold. These applications from headquarters that are now available to branch office employees could also include voice and video services delivered through WAN. On top of that, the move towards cloud computing is adding more and more strain on WAN bandwidth which needs to be properly allocated and utilized based on business and end-user needs.

The next step after you roll out full fledged intelligent branches is essentially a better allocation of WAN bandwidth, application traffic reduction, data center virtualization across enterprise, and branch office IT consolidation. Seamless application delivery to all enterprise users while being completely oblivious of their locations, technological moves towards cloud-computing architecture, WAN bandwidth reduction requirements and proper allocation of bandwidth among business needs created the market for WAN optimization products.

The problems which needed to be addressed in this space were so unique and disparate that multiple vendors surfaced and flourished. Here are some of the companies which have products to address this market place:

  • Riverbed Technology (NASDAQ:RVBD)
  • Ipanema Technologies
  • Exinda
  • Blue Coat Systems (NASDAQ:BCSI)
  • Expand Networks
  • Silver Peak Systems
  • F5 Networks (NASDAQ:FFIV)
  • Juniper Networks (NYSE:JNPR)
  • Citrix Systems (NASDAQ:CTXS)
  • Cisco Systems (NASDAQ:CSCO)

Most of these vendors started with different approaches to address these particular bandwidth management and application delivery business needs on WAN links. Juniper Networks acquired Peribit networks and Citrix acquired Netscaler to enter this market. A lot of innovation was required in this space to address a multitude of problems which needed to be solved for different network systems and application vendors. Hence many startups came along with specific niches or ways of addressing this market.

Some players approached the market by addressing application-specific needs over WAN links by handling business critical applications data and their time-out management efficiently, thereby delaying e-mail, backups and personal Web access types of data using temporary storage mechanisms on branch office servers. Data that is requested again and again was compressed or cached at the branch office for these types of solutions. Other vendors addressed this market using a network centric approach by providing better QoS mechanism for branch office data by providing high priority to latency-sensitive protocols and low priority to slow file transfers and content distribution etc. Application specific intelligence and special handling of certain protocols such as SSL acceleration for secure web access was built into the network devices itself for this type of solutions.

However, now the market is maturing and there are clear approaches which are winning. Most vendors have already integrated network based logic with application based logic in their devices. Riverbed seems to be emerging as a leader in the market along with Blue Coat Systems and Expand Networks. Exinda is still a very small player.

While analyzing this market one needs to look at developments in markets which are closely attached to this space such as virtualization and data center infrastructure consolidation i.e. Blade Servers, software and application delivery methods, cloud computing, WAN speed and traffic management technologies. The data storage systems market is going through a consolidation phase and new companies with cloud based methods of application delivery are forcing old IT implementation and services vendors out of the market. Considering the overall size prediction of the WAN optimization market is around $1 - 1.5B range in the next two to five years along with presence of both Cisco and Juniper in this space, it could easily be predicted that there would be significant consolidation in this market sooner or later.

There were some recent talks on a couple of websites saying Cisco is trying to buy Riverbed. Overall, neither Cisco nor Juniper is as successful in this market as they would like to be. Cisco, Juniper, Citrix and F5 are big players in this market. Blue Coat and Riverbed are other publicly traded companies. Ipanema, Expand and Silver Peak are privately held companies. Ipanema is a Paris, France based company and Expand is a New Jersey based company, while other companies have their WAN optimization businesses in Silicon Valley. Riverbed is clearly emerging as a leader in this market. Based on Riverbed’s current market cap and the aggression it has shown in capturing market share in this space, along with demonstration of quick innovation capabilities, it is not clear whether it will be a consolidator or target in this market.

If the consolidation happens in this space, Blue Coat and Silver Peak could be targets for Cisco although Blue Coat might have slight disadvantage as Cisco usually avoids publicly traded companies. Juniper, with its Peribit acquisition, is not successful in this space so far and has very small market share. Its recent focus on the branch office market may force it to look at other companies in this space again.

Looking at the trends in application usage and network usage behavior, there will be new technological requirements from vendors in this space such as handling increased bandwidth from home workers and mobile workers i.e. individual users. There would also be product enhancement requirements from existing vendors so they can provide better architectural capabilities for server-less branch offices, CDN, encryption and QoS solutions along with merging of network centric and application centric approaches. Clearly, for these kinds of enhancements where these products would start affecting total network architecture, small vendors would need to collaborate with bigger players either in-house or in joint development efforts.

Time will tell how the market will shape up in next two years, but clearly we would not have nine to ten players going after this market but probably four to five. Riverbed at its current P/E ratio of ~157 seems too pricey over F5 at P/E of 36 and Citrix at P/E of ~41 but it is also emerging as clear winner so far. Blue Coat Systems after its Packeteer acquisition can pose significant challenge to Riverbed and become a target for one of the bigger companies. It would be interesting to see weather Cisco, Riverbed or Juniper would make first move towards consolidation.

Disclosure: Long CSCO