Gartner Analysts Nathan Hill and Michael Silver recently published a research note titled “Define and Prioritize the User Experience to Succeed with VDI and DaaS.” The article is aimed at someone preparing to adopt or refresh a remote desktop solution, and calls out some critical learnings that I’ve also noticed in the years I’ve been in the industry.
- Delivering a user experience that’s as good as or better than an employee’s current PC experience is the most important critical success factor when adopting a virtual desktop infrastructure or desktop as a service
- Address users’ non-session-based usage requirements to ensure that they’re aligned with business and IT goals – don’t reduce the quality of the UX to save on costs
- Through 2020, more than 80% of organizations that don’t prioritize the user experience (UX) in the planning phase will fail to meet their virtual desktop infrastructure/desktop as a service (VDI/DaaS) deployment goals.
VDI UX: get it right the first time
If you wanted to pick a single aspect of a VDI or Destop as a Service project that is most likely to bring success or failure, it’s nailing the user experience. I’ve said the same thing in dozens of conference presentations and customer meetings – if it’s not as good as their PC, users won’t want to use their new VDI. The article warns further that you really do have to get it right the first time, as “it may be difficult to revive a troubled DaaS deployment, once it has a bad reputation” – and that’s the challenge. How do you deploy a high-quality user experience, if you don’t let users try it, without risking the reputation of your VDI? Fortunately, we solved this a few years ago, and it’s become the de facto standard for VDI user experience benchmarking: instead of exposing your real users to VDI, we use virtual users.
Virtual users to the rescue!
Virtual users are great. They don’t collect a salary, they don’t go on vacation, they’ll attempt to do the same tasks over and over when we ask them to, and they never complain. And they measure exactly how long it takes them to perform those tasks so we can really understand the performance. And because I don’t need to find a place for all of them to sit, it’s much easier for me to scale up my testing, or really push it to find out my max capacity.
That’s what Login VSI does – you say “show me how my system performs with 200 users doing typical office worker tasks” – and VSI will start logging in 200 virtual users into your environment, and those users will automatically start sending email, browsing the web, creating documents – all typical office worker tasks – and while they are doing them, the virtual users are logging how long the tasks take – how long did it take to copy a file? How long did it take to launch Word? As more and more users start logging on and load on VSI increases, we typically start to see those tasks take longer and longer to complete. VSI identifies for you the sweet spot – what was the maximum number of users that were running and still reporting acceptable performance. We call that your VSImax. There are all kinds of test scenarios you can run with Login VSI. Test like “what happens when 500 users all try and login at 9am on Monday morning?” or “how does my storage perform if I have 1,000 users running continuously for 168 hours straight?”
What about once you’re in production?
Once you are up and running with VDI and real users, you still want to make sure that your performance matches what you expected from your testing, and we have another virtual user solution for that – Login PI. Login PI uses the same idea of virtual users performing tasks and measuring them, but instead of having 100’s or 1,000’s of users trying to stress test the environment, with Login PI you just have one or a few virtual users logging in to VDI throughout the day along with your real users. The Login PI virtual users are watching for performance anomalies – was my login time longer than it was yesterday or last week? Did the applications launch as expected? – and if anything is out of the ordinary, Login PI sends alerts to notify you that something’s not right. It’s a great way to validate user complaints – is it really slow, or is performance on par with expectations? It also gives you notification of outages so you can start remediation before the helpdesk phones start ringing.
Don’t compromise on user experience
The Gartner article closes with this:
Don’t compromise UX for budgetary reasons, because the UX cannot be traded against to reduce costs. Instead, review whether your decision to deploy VDI or DaaS (or to continue using VDI or DaaS and address existing UX failings) will result in optimal UX. For many organizations, this will not be a question of either/or. Instead, it will be a matter of where to target these solutions across the business, when investing in a UX-optimized solution can be justified against all UX perspectives. Virtual users will help you understand your VDI UX without having real user risk loss of productivity. Login VSI and Login PI enable you to really understand how your environment performs so you can deploy with confidence.