Legacy Paradigm – Bucketing Users
Ever since the earliest multi-user remote Windows solutions of the 1990s, we’ve classified users based on broad commonalities. Sometimes this was departmental—legal users versus accounting users versus engineering users. Other times we broke users down by their assumed sophistication level—power users, knowledge worker, task workers, etc. While this wasn’t ideal, the “bucketing” of users was necessary due to the technical limitations of the day.
For example, static base images for Terminal Servers and VDI users meant that we couldn’t dynamically install applications at login time, so we were forced to create separate images for each group of installed applications, hence the “legal” bucket of users which got the legal image which included all the apps the legal department used, the marketing image which had the marketing apps, etc. We were able to make some broad assumptions about the resources that each type of user needed too: marketing needed GPUs and used multiple monitors, engineering needed more CPU and memory, legal needed special printers, and etc.
In addition to bucketing users by department, we also bucketed them based on how intense we thought they would be. (How many of us remember talking about the number of “task workers” versus “knowledge workers” when figuring out Terminal Server or VDI density?)
While that bucketing made the application selection in our base images and server sizing easier, it never truly represented our users. It was nothing more than a low-resolution approximation. For example, not every marketing department user was an actual designer who used graphical apps. Some legal users reviewed patents and needed access to CAD apps. Many administrative assistants were much more “power users” than the executives they supported who used nothing more than Office and a browser. But of course, we used those generic and broad buckets because we didn’t have the technology to build separate images or to size separate sessions or VMs for each unique user.
New Paradigm – Supporting Individual Users
This is not the case in 2018. Today’s technology allows us to package each Windows application (or groups of related Windows apps) into packages we can dynamically attach to a generic base image at runtime. We can also dynamically load-balance VMs across physical hosts and watch performance to adjust resources (up or down) based on what users actually need.
This means that the days of trying to force a diverse set of users into a narrow selection of buckets is behind us. We can support whatever resources and applications each user needs.
This extends beyond VDI and RDSH use cases. Modern EUC platforms such as VMware Workspace ONE support virtual and physical desktops, Windows, macOS, and Chrome OS; iOS, Android; and desktop, mobile, and SaaS apps. This means that we don’t have to force users into a certain device, application architecture, or desktop location. It also means we can support all the different “slices” of users within each platform. For example, “Windows-based physical laptop users” could mean ultralights, Microsoft Surface tablets, portable workstations, or Alienware employee-owned laptops. “Home users” could mean people working from home, coffee shops, or cars, using laptops, tablets, or phones. “Office workers” could be users using corporate-owned devices in the office, or users who brought in their own devices, working from their desks, or the kitchen, or a conference room, on either the corporate Wi-Fi or the guest network.
And it goes without saying that the “employee-to-device” ratio is probably one to three or four. (Every user has a primary computer and a phone, and most likely a secondary computer and maybe a tablet.) So really every employee will probably fall into more than one of the scenarios outlined above.
(Actually, this all applies to employees, temp workers, contractors, and guests.)
So, if we were still thinking in terms of the “buckets” of the 1990s and 2000s, you can see that a company with 1,000 employees would probably have to deal with two or three thousand buckets. In 2018, we no longer have the luxury of saying, “Here are the five template use cases we’ll support. Which one are you?” Instead we need to support all buckets of all users, regardless of the devices they choose, the apps they use, where they work, or when they work. Fortunately, VMware Workspace ONE makes that possible.