Why it's finally time for Microsoft to buy Citrix. Here's our full analysis.

There's been talk for years that Microsoft should buy Citrix. (2008, 2011) In the '90s, this was always counted with, "Why buy the cow when you get the milk for free?

There’s been talk for years that Microsoft should buy Citrix. (2008, 2011) In the ‘90s, this was always counted with, “Why buy the cow when you get the milk for free?” Certainly that was the case in the late ‘90s and 2000s. Citrix’s main / only product was an enhancement to Microsoft Terminal Server, and every copy of MetaFrame / Presentation Server / XenApp pulled Microsoft TS / RDSH CALs along with it. Many speculated that Microsoft’s RDSH CAL business was $1B a year—all “free” for Microsoft with them doing very little work to get it. (We joked that revenue per person in that group was off the charts! $1B from maybe 35 folks on the team? That’s $28M a head!)

Of course parts of this are still true today. Microsoft RDSH CALs (or SA) are required for every XenApp / XenDesktop deployment, and the more Citrix sells, the more Microsoft gets. (This applies to Horizon CALs too.)

But in 2015, the battle is not the Windows CAL. The battle is the datacenter infrastructure. It’s all about "Hyper-V versus vSphere."

This is slightly awkward for Microsoft since Citrix XenDesktop and XenApp can run on XenServer, Hyper-V, or vSphere, and today the biggest share of these Citrix environments run on vSphere. (So the Windows part of Microsoft wins, but the datacenter platform part of Microsoft loses.)


VMware Horizon only runs on vSphere. With their recent doubling-down on the desktop, VMware is focusing on a true end-to-end story, from datacenter virtualization infrastructure all the way to the desktop and mobile device.

In 2015, Horizon about extending VMware’s datacenter platform out to the end user. Horizon is an extension of vSphere. It’s not a desktop virtualization product. To this end, when you talk to VMware employees about the desktop virtualization market, they don’t focus on Citrix as a competitor—they focus on Microsoft. Everything they’re doing across the company—including around Horizon—is geared to beat Microsoft in the datacenter.


Microsoft is obviously serious about the datacenter and owning the virtualization infrastructure at customers. In this regard, their competition is VMware. So we have two big companies pointing their guns at each other. (When you talk about the larger cloud, you include companies like Google and AWS, but when it comes to the datacenter infrastructure that runs enterprises today, that’s a pure Microsoft and VMware thing.)

So Microsoft views VMware as their competitor here, and they see VMware doing and saying all the right things—including all the right things about end user computing and Windows desktop and application virtualization. As I wrote last week, I not believe that Dell’s pending acquisition of EMC will hurt VMware. I’m not sure whether I can say it will make them stronger (my sense is “yes”), but I feel certain that it won’t hurt them. (Though really just getting Elliott off their back can be viewed as a strengthening factor of this potential deal.)

This means that from Microsoft’s perspective, their main enemy is doing and the right things and getting stronger, and they have a complete end-to-end story to sell from the datacenter to the desktop.

What’s Microsoft’s counter-argument today? They have Hyper-V. They have Azure. They have Windows Server. They have InTune and System Center. But to really compete in the Windows application and desktop delivery space and to really “own” the end-to-end, datacenter-to-device story, Microsoft is missing key functionality that they rely on Citrix for.


Microsoft relying on Citrix worked well for the past 15 years. But that’s all changing.

Just as VMware is getting stronger, Citrix is getting weaker. We’ve written quite a bit about the recent struggles at Citrix. (here, here, here, here, here, here, here, and here) They’re getting pressure from Elliott. Employees leaving. Negative quarter-over-quarter sales. Looking for a new CEO. Product and sales execution issues. Questions about strategy. The list goes on and on . . .

By now Microsoft must have realized that they can’t rely on Citrix to deliver a critical piece of their end-to-end datacenter-to-desktop strategy. A weak Citrix is very bad for Microsoft. (And remember, most Citrix deployments are on vSphere, so that just helps VMware get stronger and provides customers with another reason *not* to move to Hyper-V.)

When you talk to enterprise customers, you learn real quick that there are “Hyper-V shops” and “VMware shops.” The “VMware shops” have a complete end-to-end solution from VMware. The Microsoft shops have to buy from Microsoft and an ever-weakening Citrix, (and the Citrix part can run on VMware!). This is not good for Microsoft.

Microsoft buying Citrix

When I wrote about the Dell/EMC deal last week, BrianMadden.com reader vgernyc wrote that it makes sense to split Citrix between Microsoft and Cisco. I agree with this.

Let’s run down Citrix’s products and where they might most make sense if Microsoft were to buy them:

XenApp / XenDesktop

XenApp and XenDesktop can go to Microsoft and become standard server products (like Exchange and SQL). Microsoft can bundle them, build core features into Windows Server, offer them as part of Azure, and/or sell them as standalone products.

Just as VMware has a “hook” into customers who use VMware infrastructure, Microsoft would have an equally-powerful “hook” into customer since they use desktop virtualization to deliver Windows desktops and applications. In fact Microsoft would be able to offer a true end-to-end solution since the infrastructure, the Server and Client OS, and the delivery could all be from Microsoft. VMware could own the infrastructure and the desktop and application delivery for their customers, but there would always be that middleware Windows layer which was still owned by Microsoft. (And it will be for a long time since I don’t see Windows apps going away in the enterprise anytime soon.)

In fact Microsoft could really up the pressure by extending the core Windows Server offering to include all the functionality of XenApp and XenDesktop—essentially making them “free” and forcing customers to decide whether they need to pay VMware for Horizon when they can get it for free from Microsoft. (Not that that would be unfamiliar territory for VMware since that’s essentially what Microsoft does today with Hyper-V versus vSphere.)

Citrix Workspace Cloud

I’ve written and said many times (here, here) that Citrix Workspace Cloud (CWC) is a fantastic product and represents the future of XenApp and XenDesktop. I love the concept, I love the technology, and I love how well Citrix has executed well on it. (The only stumble is the pricing which seems out of whack to me. If you subtract out the included costs of the core product licenses, it seems like Citrix is trying to charge $10-15 per user per month just to run some SQL and web servers for you?!?)

From a technology standpoint, CWC is where you pay for your Citrix products on a per-user, per-month basis, and then Citrix runs your configuration database and web servers for you. The core XenApp / XenDesktop and data stay on servers in a location of your choosing (whether in your datacenter, in a cloud provider’s datacenter, or a combination of both).

I assume that the services that Citrix hosts for CWC customers already run in Azure (even if the resource servers a customer chooses are elsewhere), so this should be a simple thing for Microsoft to take over. Microsoft could also adjust the price so it was more in-line with what people expect to pay—perhaps something like $2 per user, per month for all the back-end infrastructure as a service needed to enable users to connect to their Windows desktops, apps, data, and configurations from anywhere (again whether they’re on-premises, in the cloud, or both).

Microsoft could also use this to create a full-on DaaS offering (to complement their limited Azure RemoteApp) which could compete against VMware’s Horizon DaaS, and Microsoft could give customers the added benefit of flexibility to run their actual resource servers wherever they want. (Something VMware has talked about but does not currently offer.)


I called Jack and asked him about XenMobile and what it might mean to Microsoft, so everything I’m typing in this section is based on what he said.

First, Microsoft already has a XenMobile competitor with InTune and the Microsoft Enterprise Mobility Suite (EMS). Brad Anderson has written about how Microsoft’s approach to EMM is different from others in that it was home-grown and built for the cloud. (Part of their “mobile-first, cloud-first” strategy.) So it’s not like Microsoft needs to go shopping for features.

That said, there’s not a lot of independent data about what the real-world adoption is. Microsoft has talked about having over 10k customer using it, but it’s hard to tell what actually counts as a “customer” since Office365 includes native mobility management capabilities which means they’re most certainly counting every O365 customer as an EMM customer.

That said, XenMobile is a legit product with a measurable market share, so worst case this could be a classic “one competitor buying another” and then rolling all those new customers into their own product. This could be slightly challenging since XenMobile offers an on-premises solution (actually that’s the main thing people use) whereas everything Microsoft is doing is cloud-based.

So it’s unclear if Microsoft would go down the path of creating an on-prem version or whether they’ll just force everyone to migrate. (I would suspect after reading Brad’s blogs that Microsoft would not go down the path of offering an on-prem version.)

On the integration side, Citrix and Microsoft have both been building lots of mobile apps to enhance their EMM offerings, though customers would probably prefer Microsoft’s versions since with Microsoft you can get the “real” versions of Office and along with the assumption that they’ll have the best integration with the rest of Microsoft’s application stack.

The reality is that Citrix and Microsoft compete pretty much head-to-head in the EMM space, so Microsoft buying market share while simultaneously removing a competitor would be good for them.

Citrix Workspace Suite / Portal

One thing that Citrix has done for years (and that VMware is doing now) is offering a single portal users can go to in order to access all their apps and data regardless of whether they’re looking for a Windows, web, mobile, or SaaS app, regardless of where the data is, and regardless of what type of device they’re connecting from. (And with intelligence to deliver the right technology based on the user’s device, location, profile, security, etc.

So Citrix and VMware have been going down this path for awhile, but so far this hasn’t been something we’ve seen from Microsoft. (There’s no “Microsoft Workspace” product or anything like that.) Sure, there’s SharePoint (which started life 15 years ago as “SharePoint Portal Server”), but that’s more about web-based collaboration and document management.

That said, SharePoint does allow apps to plug-in to it (and Citrix has long offered plugins for published apps and desktops), so I could see a nice fit for Citrix’s workplace vision to be more tightly integrated to SharePoint (and the related apps) to provide integrated, location- and client-aware access to legacy applications and desktops alongside everything SharePoint offers today.

This could also fit nicely into the various other offerings Microsoft is building for Azure, Azure Active Directory Premium and some of the Windows 10 profile management stuff. You can envision that being able to dynamically assemble the right mix of apps with the right technology is something that would be great for Microsoft and something they don’t really have now. 

Citrix Online (The “GoTo” products)

There was talk about Citrix selling off their “GoTo” products entirely, but it seems that all the core GoTo offerings (GoToMyPC, GoToMeeting, GoToAssist, GoToTraining, GoToWebinar, GoToWebcast) would fit well within Microsoft’s growing service offerings with very little overlap. I would imagine that Microsoft would want to keep them and roll them into everything else they offer. Some might become features of Skype, Lync, or Office365, but in general it seems that every Citrix Online product would be a good product for Microsoft to own for both strategic and practical reasons.

I also like the idea of the clients for these apps being integrated into all the other clients for the other Microsoft apps. Office365 and GoToMeeting integration, launching ad-hoc GoToMeetings from Office365 apps, integrating Office 2016 real-time co-authoring, etc.


While ShareFile isn’t super popular among consumers, many enterprises like it since it lets them run their own Dropbox-like service with on-premises data storage. I would imagine this could fit nicely into Microsoft OneDrive and SharePoint, ultimately providing key technology components that could enhance Microsoft’s OneDrive for Business on-premises offerings. Worst case they get a bunch of new customers for OneDrive with another classic “one competitor buys another” scenario.


NetScaler is really the only component that doesn’t fit well into Microsoft, so as commenter vgernyc pointed out, it probably makes sense for them to sell this business and I could see it getting picked up by someone like Cisco.

The only complexity is that Citrix spent a lot of time over the years building XenApp and XenDesktop features that require NetScaler. I’ve written many times that I wish Citrix would open up XenApp and XenDesktop so that they can work with any application delivery controller (ADC), and certainly Microsoft would probably do that, but in the meantime they would need to ensure that a Microsoft XenDesktop and XenApp would continue to work well with whomever bought this.

Or maybe Microsoft keeps it and focuses on the virtual networking components which they can build into Hyper-V?


I’m not sure this is even a real product anymore, but any IP Citrix has that’s interesting here can just get rolled into Hyper-V and/or Azure. Worst case I would think that any remaining XenServer enterprise customers would be more likely to move to Hyper-V over VMware if Microsoft absorbed the product and rolled it into Hyper-V.


Citrix doesn’t have a DaaS offering per se, though they have lots of extensions and tools to help hosting providers host Windows desktops and apps with XenDesktop and XenApp. All of that could be rolled into Hyper-V and/or Azure and can be used to help manage desktop and application hosting on Azure or any other cloud platform.


I know Octoblu is not exactly a real product within Citrix, but since they spent so much time talking about it at Synergy I figure I’ll include it here. :)

I’m still not 100% sure what Octoblu does from the enterprise standpoint, but Microsoft is also making a lot of noise in the Internet of Things space, so I would think Octoblu can snap right in to what they’re doing there.

Is a deal realistic?

To assess whether this deal could happen, you have to look at both (1) whether Microsoft can afford Citrix, and (2) whether it strategically makes sense for them.

From the cost standpoint, Citrix has a market cap of $12.2B. If Microsoft pays a 7% premium, figure that’s a price tag of $13B. But they will also raise a few billion by selling NetScaler, so that lowers the cost a bit.

Microsoft has $100B in cash (albeit $95B of that is outside the US), so it’s not like this is an impossible transaction for them.

From the strategic standpoint, I’ve made the case in this article, but I’l summarize with this. Microsoft’s competition in the enterprise is VMware. VMware’s “best case scenario” is for Citrix to continue to exist as a standalone company. Certainly Microsoft doesn’t want to sit back and watch Citrix nose-dive while simultaneously handing VMware a silver platter with their best-case scenario.

Given Citrix’s current situation, I can’t see them staying on their own too long, so now we’re looking at the possibility of Microsoft or someone else buying them. At the end of the day, I can’t see Microsoft wanting to risk someone else (HP? Cisco?) picking up Citrix which would put Microsoft right back where they are today having to rely on a third party to give them the complete enterprise solution.

(Maybe instead of buying Citrix outright, Citrix will be able to raise the cash to go private and Microsoft could throw in a few billion to get a piece of them—and “own them without owning them” sort of thing?)

I’ve always said that Microsoft will buy Citrix when they have to. (And not a day sooner!) It looks like that day is near.

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

I really can't see the Citrix / Microsoft thing. They're competing on too many fronts for this to be mutually beneficial. Plus the Microsoft + Unidesk love fest at the moment is my next acquisition guess.

Cisco on the other hand makes a lot more sense. Xenserver, netscaler, Melio if it ever materialises in dom0, etc would give them life again if VCE goes the way of the dodo. The workspace technologies (XenMobile, XenDesktop/App) would add new business and Webex is pants anyway, so the Goto suite would be a better product.

It is best for Microsoft to buy Citrix and sell it products as the bundled package with its existing products like system center.