Dan's Take

How Citrix Got Here

Citrix has taken radical new steps to better compete. Here's why it began to struggle.

• By Dan Kusnetzky
• 11/20/2015

Citrix has just taken steps it believes will help it adapt to a changing world of IT, spinning off the GoTo products and laying off staff. I'd like to examine some of the reasons behind these moves.

Citrix Origins
Citrix originally offered access virtualization technology (see The 7-layer Virtualization Model for more information about the layers of virtualization technology) as its primary focus. Citrix's technology made it possible for applications executing on Windows and Solaris servers to be accessed by users of OS/2 and Windows client systems as if those applications were running locally. Over time, OS/2 clients were dropped, as well as Solaris servers. Later, Citrix focused its efforts on making it possible for Windows applications to be accessed by users of many varieties of smartphones, tablets, laptops and PCs.

Microsoft has long offered its own access virtualization technology. As that technology is integrated more deeply into Microsoft's applications and management tools, Citrix has faced the challenge of maintaining its customer base.

As enterprise workloads have moved to be either Web- or cloud-centric, virtualizing access to Windows applications is of less importance. Access to applications and data in this emerging world is based on the use of Web browsers or application-specific apps. An access virtualization layer isn't really needed, but Citrix has been slow to address the erosion of its initial business.

Covering the Virtualization Waterfront
Over the years, Citrix has developed or acquired technology that allowed it to compete in the market for application virtualization, processing virtualization, network virtualization and, most recently, storage virtualization.

Although it had entries in each of those segments of the virtualization market, the company had to walk a very fine line due to its strong partnerships with other suppliers of that type of technology (chiefly Microsoft and VMware); it worried that too much overlap could sour those relationships.

That desire to "still be friends" with its competitors had a dampening effect on Citrix's efforts to answer the challenge by Microsoft and VMware.  Instead of trying to be the primary source of virtualization technology for its customers, it seemed content instead to collaborate.

It was the wrong strategy. As enterprises increasingly relied on VMware, Microsoft and the open source community for virtualization technology, Citrix found it difficult to capture new users with its offerings.

The Battle for Xen
Digging a bit deeper into the processing virtualization segment of the overall virtualization market, it's clear that Citrix has been behind the power curve for quite some time. In its early efforts to become known as a virtualization company, it acquired XenSource, a company that was founded by the original developers of the Xen hypervisor. At the time, many industry analysts, including my former employer, IDC, thought that Citrix paid too much for XenSource. The result was that Citrix had to make its Xen distribution costly when compared to software offered by others. Citrix found it difficult to charge high fees to use technology that was then, and still is, a common component of every Linux distribution. This meant that customers didn't flock to Citrix's version of Xen; they simply went to suppliers such as Red Hat and SUSE.

Citrix was dealt an additional blow when another of the company's partners, Oracle, brought out its own Linux distribution, along with its own distribution of the Xen hypervisor. This made it difficult for Citrix to win over Oracle users.

Xen, however, isn't the only virtual machine (VM) software technology in the market. The availability of other hypervisor technology has had its impact on Citrix as well.

Hypervisor Wars
While the Xen community was battling with itself, hypervisors from VMware and Microsoft were claiming more market share. VMware suggested to customers that its technology was more mature, robust and better supported. Microsoft pointed out that Hyper-V was more tightly integrated into the Microsoft environment and could be supported more easily by Microsoft's management tools. This meant that Citrix had to compete strongly with its good partners to win the hypervisor.

The open source community, meanwhile, began to adopt the KVM hypervisor. Many suppliers, such as IBM and Red Hat, made it the default rather than Xen. While they continued to offer support for Xen, it was no longer their primary hypervisor. Since Citrix didn't offer strong support for KVM, it was typically cut out of those environments.

Behind in Containers
The Kusnetzky Group model recognizes five different types of processing virtualization technology. VM software is just one of the five, not always the best choice for certain applications. Operating system virtualization and partitioning (OSVP), in the form of containers, has gotten a lot of attention because the computing environment can be more efficient. OSVP doesn't require each virtual workload to have its own operating system and all of the memory, processing power and storage it requires. Although Citrix has spoken about its support of containerized applications, it isn't seen as a leader in this exploding movement, another weakness.

Other Forms of Virtualization
Network and storage virtualization are also hotly contested areas in the market. Citrix has long offered NetScaler to address network virtualization, but it too has been overshadowed by other suppliers of networking technology.

It was the same story with storage virtualization; late to the party. Citrix hoped its acquisition of Sanbolic would help it leapfrog the competition, but that hasn't happened either. Although its offerings in these areas are good, they've gotten somewhat lost amongst Citrix's many messages.

Cloud Computing
Although Citrix was one of the early contributors to OpenStack, it decided to focus more on CloudStack. Several industry observers believe this is because it could more closely control CloudStack, and OpenStack was being driven by a broad set of industry players and customers. Rather than follow Citrix into the world of CloudStack, most of the industry continued to focus on cloud infrastructure from Amazon, Microsoft and the OpenStack community.

Online Collaboration
Citrix's online collaboration products, sold under the "GoTo" name, have been very successful, but haven't been able to pull the company out of the doldrums. Why? There are many other alternatives from suppliers such as Adobe, Microsoft and the open source community.

Dan's Take: To Survive, Citrix Must Get Ahead of the Curve
The world has changed much faster than Citrix has. Its core business, access virtualization, is of less interest in the world of the Internet of Things. Its revenues coming from virtualization technology have been surpassed by those captured by suppliers such as VMware and Microsoft.

Its primary processing virtualization technology, Xen, isn't a serious player. Even if an enterprise was interested in using Xen, it's likely that it would select lower-cost options.

Spinning off the GoTo family of products might allow Citrix to focus more on its stated mission of making it easier to access enterprise applications. That being said, the company is being challenged on many levels. Energetic competitors at each layer of the virtualization model have nibbled away at Citrix's mindshare. Former partners are increasingly focused on selling their own technology rather than going forward with Citrix to win new customers. Enterprise developers are increasingly relying on browsers or special-purpose apps, rather than access virtualization technology.

Citrix has good people and good technology. It needs to focus on getting ahead of the wave of technology, rather than on catching up with what others are doing.