Cisco’s SDN Challenge

In “The Godfather, Part II”, Michael Corleone famously said “keep your friends close but keep your enemies closer.” When it comes to Software Defined Networking (SDN), Cisco is following the Godfather’s advice.  SDN is viewed as a technology that will fundamentally change how networking is approached, and as the leading networking products vendor, Cisco has more to lose than anyone else if SDN is successful and has more to gain than anyone else if it is not. While they have an aggressive plan to embrace SDN, they also have the best motivation for maintaining the status quo.  How this plays out in the end depends on how SDN fares in the market overall and how deeply Cisco is actually willing to dive into providing solutions.

Networking today is characterized by multiple tiers of switches and routers that distribute information to servers and client devices.  The core routers sit at the heart of the data center and have the main control over the distribution of information. In the networking hierarchy data moves from the core to other “leaf switches”, first at the end of every row (EOR) of racks in the data center, then on to the switches at the top of each of the racks (TOR). Cisco owns roughly 65% of the core routing market, more than double the next closest competitor, Juniper.  The core is the most profitable business, but the EOR and TOR businesses also contribute to the business, not only from a profit perspective, but more importantly from a volume and revenue perspective.  Control the core and you have better control over what happens at the row and rack level.

Earlier this year Cisco announced its Cisco Open Network Environment (Cisco ONE) and the juxtaposition of the word “open” in a Cisco announcement has caused plenty of head scratching.  Cisco prides itself on its technologies and is not necessarily one to push an open source standard; their business model is about captive customers and sunk investments that create a barrier for customers who want to switch to competing technologies.  This is not to say that Cisco does not have great products; they would not be where they are today without great technology, but, it does say that they have an incentive to keep things proprietary. When Cisco claimed that they were introducing “the industry’s most extensible open networking fabric,” that news was met with some skepticism.

Cisco ONE is a very broad strategy that encompasses a wide range of networking functions. Obviously they are staking their claim in SDN and trying to not fall into the same mistakes that they have made in the past of leaving areas of value uncovered that allowed companies like F5 or Riverbed to carve out a niche for themselves.

As Cisco tries to bring this strategy together, they have a huge challenge ahead of them.  With 3 different operating systems (IOS, IOS-XR and NX-OS) that are all differing technologies that have different release cycle cadences, the software that Cisco is eyeing to bring to the market will need to comprehend all of these and somehow tie the release cycles together – or be constantly updating, which is not acceptable to customers who relish stability and longevity in releases.

In understanding how Cisco will fare in the SDN market, it is easiest to dissect the situation and break out a standard SWOT analysis to look at their Strengths, Weaknesses, Opportunities and Threats.

Strengths

Weaknesses

  • Market Share
  • Ecosystem
  • Channel
  • Profit
  • Market share
  • Over-featured
  • Customer perception
  • Overly attractive margin pools
  • Saturation of skilled professionals
  • Innovator’s dilemma
  • Over-featured
  • Customer perception
  • Overly attractive margin pools
  • Saturation of skilled professionals

Opportunities

Threats

  • Army of professionals
  • Truly embrace open source
  • Outflank the completion in TOR/EOR
  • Expand certification
  • TOR functionality moves out of switches
  • VMware
  • Focus on software over hardware
  • TOR functionality moves out of switches
  • VMware
  • Focus on software over hardware

 

Strengths

To begin with, Cisco’s major strength is their market share.  In the networking world, customers are loathe to switch vendors, especially if everything is running well.  They’d rather spend a little more (on in Cisco’s case, often a lot more) in order to “buy” the stability and consistency of a trouble-free networking platform.  With the market presence that Cisco commands today, their inertia helps them maintain share and profitability. High share today translates into high share in the future as the rate of defection to other vendors is fairly low. In this stable market Cisco can play the king and extract a healthy margin for their business.  The switching costs help maintain this business and keep profits high.

Also, there is a large base of Cisco administrators and developers.  This piece of the ecosystem helps them maintain their presence.  Cisco certifications like CCNA and CCNP carry a tremendous weight for the holders of those letters when they apply for jobs.  Like a trade union, having the certification means higher salary, better jobs and a more stable future. Their top-notch training programs churn out an effective number of new graduates each year to keep pace with the market and the rigorous testing programs help ensure that those letters mean a lot to potential employers. All of this feeds back into Cisco being a preferred brand that holds its values.

The Cisco channel partners, as well, are a strong asset for helping move product out from Cisco to the end customer.  With the new focus on SDN this group of resellers and VARs will be called into active duty to help drive the message out and place the products into the hands of customers.

The business, as we have said before, is very profitable so both the revenue and margin are strong. As a company, Cisco is investing more back into maintaining its market share, not necessarily milking the market.  With gross profits north of the 60% mark, almost double some competitors (ZTE, Alcatel-Lucent, etc.) there is definitely a strong backing to keep that business rolling. Chief amongst the financial factors is one that is tied more to their customers than their company:  the majority of the large, core routers that are installed today have not been fully amortized. This means that even if a company was thinking about switching, there is not currently the financial incentive to take action yet.  This buys Cisco time to see how SDN shakes out.

Weakness

When examining weakness for a company it is important to note that almost every strength can also be viewed as a weakness to some degree.

While market share may be a strength for Cisco, it also poses a weakness as Cisco is reliant on that revenue stream to fund its business and any deviation from the current market trends could spell trouble for the manufacturer. They have built a successful business around driving Cisco routers and switches to all levels of the networking hierarchy and providing a set of management tools to keep all levels running smoothly.  But SDN threatens this with the potential to both flatten the hierarchy and also remove some of the functionality out of the lower level switching.  In this world, it is the market share leader who has the most to lose as market disruption changes how people deploy. Defending a leadership position is not always enviable when forces are changing the market where you lead.

In an environment where the TOR or even EOR switches no longer need the deep set of features, Cisco may find that its products are over featured for an SDN world and this will push them out of opportunities where customers desire a lower-cost, lower featured product because their SDN controller is centralizing the functions that were previously being handled by the leaf switches.

The general customer perception is that Cisco makes great products, but those products are very expensive and customers feel locked in to a great degree.  With every earnings call customers are reminded that they are paying a premium to install Cisco kit and most believe that, while they are getting good equipment, they overpay for it.  This is evident when competitive threats push Cisco, on a deal by deal basis, to offer discounts of up to 80%.  When this happens customers recognize the real impact of a Cisco environment and it underscores the concern.  Customers are interested in things that will give them more control over their environment and this is why SDN is so appealing to so many customers.  Just as they have seen the x86 server market commoditize with prices plummeting over the years and features/performance increasing at the same time, they look to their networking stack and wonder if SDN is the tool to help them achieve the same effect for the network.

Margins have been high for Cisco and Juniper, which makes the market opportunity attractive for competitors.  While there have been barriers to entry (training, infrastructure consistency, tools, etc.) for competitors in the past, SDN offers an opportunity to break from the tradition, and this opens up the market to other companies.  Earning tremendous returns in an era of disruptive change makes you a target, and the Cisco profit pools have a huge bull’s-eye on them as competitors will target Cisco’s customers with more aggressively priced SDN solutions.

The armies of trained professionals, both on the sales/partner side as well as on the administrative side are eyeing SDN as a clear disruption in the market.  Having a CCNA or CCNP certification is very valuable in a stable environment and allows both sides to extract more value, but in an era of disruption, when SDN potentially shifts demand, those skills might drop in value, making those who did not embrace change look more like Blockbuster Video franchisees in an era of NetFlix and iTunes.  The strength of the ecosystem can become a millstone for Cisco if they cannot figure out how to leverage these professionals and bring them into the era of SDN quickly.  With the emphasis on OpenFlow and open source in the SDN world, there may be hesitancy in professionals seeking out a Cisco-only SDN solution and the smart ones will work to parlay their existing knowledge into a broader-based set of networking skills to prevent another lock-in down the road if market disruption occurs again.

Opportunities

Where does this leave Cisco, how can they leverage the opportunity of SDN? Clearly there is an opportunity to transition some of their business over to SDN solutions and maintain customers while they try to hold on to their installed base.  This will not be an easy task but it can be accomplished.

An area to start to focus should be the developer and professional services areas where their army of professionals can be leveraged to tell the Cisco SDN story.  Their interaction and relationships with customers puts them in the trusted advisor category, allowing the message to flow through with less resistance. As the market craves education on SDN and how it can impact their business, Cisco is in a good position to leverage that opportunity by aligning its forces to proactively help the market understand the changes afoot.  However, the critical piece here is making sure that the message they take to market is not heavy-handed and instead focuses on the customers’ needs and not pushing the Cisco party line.  Customers are looking for change and they need to understand how they can change and still leverage their existing investments as long as possible.

For Cisco to truly leverage this opportunity they need to embrace and promote open source.  The idea of “open” typically flies in the face of Cisco’s direction in the market which focuses on wrapping customers in the Cisco cocoon so that they can enjoy the bliss of an all-Cisco world.  The driver behind SDN is open source (predominantly OpenFlow) and it is critical that Cisco accepts this and takes that message out to the market.  Customers are reaping the benefits of open source on their server software today and the comfort level with open source is high; failing to properly embrace open source will make Cisco appear like Microsoft in the early 2000’s when open source was challenging proprietary software.

Cisco has an opportunity to outflank the competition with an “SDN TOR switch.” Today’s current networking products have been focused on converged infrastructures, encapsulating Fibre Channel into Ethernet packets (FCoE) in order to reduce complexity.  If Cisco can shift its focus from converged environments to SDN and bring a lower cost SDN TOR switch to market, they can prevent some of the potential defection of customers. This product would allow customers with an existing Cisco investment to continue to leverage that investment as they move to an SDN world.  Think of this as an insurance policy on SDN – if it takes off you are covered, if it does not you are still wrapped in the warmth of an all-Cisco environment.  Just adding SDN to its current (read: expensive) switches will not present a compelling opportunity for customers and will open the door to other vendors to take their space.

By expanding the Cisco certification programs to focus on SDN, the company can continue to leverage all that they have done in providing education to the market and also help arm their forces with the tools that they need to simultaneously address SDN while trying to keep customers engaged.

Threats

The single biggest threat to Cisco is the advent of inexpensive TOR switches from China.  Hyperscale customers are already implementing these at a rapid pace, moving to SDN and OpenFlow, saving potentially tens of thousands per rack in the process.  With deployments that can include thousands of racks, that savings starts to stack up quickly.  Just as the Hyperscale customers have moved away from branded OEM systems, preferring to go right to the ODMs to have their own stripped-down compute engines built to their specifications, the network was an easy follow-on and now they are employing that same model to drive cost out of their networking.  This could not have been done without OpenFlow, which continued to drive the switching functionality that customers demand.  The networking market has never been this close to hardware commoditization than ever before.  And although it has been built on standards like Ethernet for the past 30 years, all of the functionality above the transport layer was anything but standard.  With SDN, that functionality outside of transport is moved to the SDN controller and the actual switching hardware becomes commoditized.  Normally the mindset was “he who controls the core controls the edge,” but with SDN, if all of the edges change rapidly, there is an opportunity for Cisco to start to lose core business because the value proposition is diminished. SDN is marginalizing the TOR/EOR market and Cisco needs to understand how to properly react to this or risk losing significant share from overseas vendors.

The real threat, however, from SDN comes from companies like VMware, not from hardware manufacturers.  Networking is about control more than it is about hardware.  VMware is one of the most credible challengers to SDN through two very beneficial pieces: Nicira and virtualization. Nicira was VMware’s $1.2B purchase of a network virtualization company. This purchase allowed VMware to add network virtualization to its portfolio in a manner far beyond the VSwitch functionality of their software which only controls layer 2 switching for VMs on a host.  With Nicira, VMware fired a strong shot into the SDN market and their strength in virtualization was a key piece of why they will be successful here.  VMware, the current leader in virtualization, is predisposed to be a leader in the network virtualization. With so many customers trusting VMware to virtualize their compute, it only makes sense for them to trust VMware to tackle the network as well. VMware hypervisors are driving more than half of the virtualization environments today, and integrating Nicira into existing VMware installations, while not a “plug and play” operation, will still be easier for customers.  It is easy to see that the first instances of network virtualization will probably be deployed in close proximity to server virtualization, so VMware stands to gain here in the initial phases.

The underlying threat to all of the market, not just Cisco, is that SDN represents a significant disruption to the market, and it is a disruption that puts software in the drivers’ seat, over hardware.  In an industry that has grown over the years viewing hardware as the differentiator and the vessel that delivers services and consulting revenue, this fundamental shift will have the most impact on the market.

Moving Forward

In light of how they are currently positioned today and where the market is heading, what can Cisco do? There are a few key areas that they need to focus on to ride through the disruption of SDN.

  1. Give customers a reason not to leave. While this may sound self-explanatory it needs to be stressed.  In times of disruption customers look to the new, the different. Networking has been entrenched over the years in a conservative, staid set of products and processes.  Customers are looking for more value and more flexibility as the rest of IT has changed (first storage virtualization, then server virtualization and finally network virtualization).  Customers are actively investigating alternatives and Cisco either needs to be part of the solution or risk being viewed as part of the problem. At this point their market position and profit clearly point them towards the latter; they need to focus on customers and the reasons WHY customers are finding SDN of interest if they want to get ahead of the curve.
  2. To be a part of that solution, they need to help customers understand how to bridge the transition.  With trillions of dollars of Cisco products installed around the world, customers are not anxious to pull these investments out just yet.  As a matter of fact, the uncertainty around SDN means that customers will be willing to hedge their bets.  Cisco stands to gain in this area if they can help customers prepare for the transition and then take it if they deem the ROI acceptable.
  3. Everything in Cisco’s portfolio needs to either be clearly messaged as a part of an SDN environment if it has that capability.  But this goes beyond just saying “OpenFlow compatible.” Customers want to know how, and more importantly why these products should be part of their strategy. Now is the time to allow customers to self-segment, they are either on the road to SDN, in the investigation phase or not taking that trip. There is enough room for products to cover all of these positions, but clarity of the product line will be critical.
  4. Strengthen the “open” message of their SDN solutions.  Calling their strategy “open” is one thing.  But the proof will be in the actual execution.  This is probably the area that customers are most skeptical as Cisco has been very proprietary up to this point.
  5. All of the Cisco business units need to be aligned and working towards the same goal. Today business units compete against each other and this leads to confusion amongst customers. Everyone needs to be on the same page.

Cisco stands to gain the most or lose the most as the market moves to SDN – their actions alone will determine whether or not they are successful.

John Fruehe is a guest blogger at Moor Insights & Strategy and was previously vice president at NextIO and director at AMD.  John also spent nearly 15 years at Dell and Compaq in enterprise product, strategy, and marketing roles.  You can find his full biography here.