Spectrum Protect 2016



I joined the team on September 2015, and on January 2016 I became the lead for all the design work for the vSphere build out work that lead to the use of scheduling, the ability to restore virtual machines within the GUI and worked with development in a third swim lane to deploy a new technology to the way IBM handles virtual backups. 


When I joined the team, I was partnered with the GUI Architect, a visual designer and one other UX designer. Once I was caught up on the research that had been done on this project, I was able to contribute immediately towards generating the initial concepts and wireframes that were shown to the team to gain alignment around the direction of the project for the year. 


As business needs changed, my team became myself, the visual designer and the architect. However with proper planning and prioritization we were able to not only get the work done that we planned at the start of 2016, but the restore features, this was something that was planned for 2017 and we were able to get the designs done and validated with enough time for development to get it ready for launch.


I was able to conduct user testing throughout the development cycle, however there were two main events where the bulk of our validation came from. The first was in March 2016 when myself and the two project architects were sent to Canada to conduct a recap of progress to 64 users that were in attendance. We were able to validate and pivot on our current design concepts we had, as well as an opportunity for continuing research for the next versions. Leading up to the second event I was able to conduct several rounds of internal user testing with developers and current tech support reps, to show them what we had planned. Then once we had live demoes to test, we reached out and were able to test that with our sponsor users. (The Sponsor User program lets users of our products to sign an NDA and then they can provide feedback to us during the development stages). 


When dealing with different project stakeholders, developers, offering management, and other designers getting everyone on the same page can be a challenge. Being on a legacy team, a third of them loved to talk tech and when they would present progress during presentations IBM calls 'Playbacks', they would always go down this technical rabbit holes and derail and disengage the other 2 groups. I was able to evangelize design thinking on my team and get them to embrace the use of story telling and using design as the common language to increase engagement and feedback through all three channels. 


So those are the highlights of what was done, but before I get into any details, allow me to frame the context of what is next with an analogy. I am going to make the assumption that neither vSphere nor IBM Spectrum Protect and what these things do can be considered common knowledge.  

If you use a MacBook, at some point you may have ran a Time Machine backup (I will adjust this analogy to Windows on request). It is a pretty simple process, you take an external hard drive, plug it in, and once you configure some initial settings the back up begins and all your data is now copied on your hard drive. Every time you plug in the hard drive, it makes a copy of any new changes and will continue to do so until you run out of space and you would need a new hard drive.

Pretend for a moment there is a giant call center with 10,000 employees that would each need their own MacBook. Then you would need a substantial staff to deploy the proper software, run updates, and back them up. What vSphere does is allow a company to use much cheaper hardware but you sign into a much more powerful machine that is 'virtualized' and the employee can not tell the difference. These virtual machines can be created, updated and maintained in bulk and only a few administrators are needed to maintain a massive environment. However vSphere does not handle backups and that is where Spectrum Protect comes in, we are the Time Machine backup. 

The customers of our products are changing and they are demanding a GUI based solution that is easy for them to use?

Respecting the NDA here, essentially there was an old way of doing things in IBM that was really effective, however it was equally cumbersome. The new people doing these jobs, prefer GUI's over writing Powershell commands. The new method of protecting these virtualized environments involves creating a schedule that would tell a group of virtual machines when they are allowed to backup to IBM. 

Which created the following questions:

  1. How do you set / edit a schedule?
  2. Where are schedules created? 
  3. How do you monitor / fix things? 


Problem 1: How do you set / edit a schedule?

Information Architecture

Integration with vSphere was one of the attractive value additions to this solution, however vSphere has its own rules and design guidelines that we had to follow. So this was the map we used to communicate across the design team, and to determine the touch points of where our solution would eventually go.

 Everything in white are the existing vSphere panels and everything in blue is where the IBM Data Protection would integrate. 

Manage Panel (Image)

vSphere a series of fixed header tabs where we were able to implement our design. Here the user can select a schedule that is provided to them and there are a series of options that through user testing are presented to the user for editing. The initial wireframes allowed for the schedules to be created and edited under the manage tab

Container Manage.png

Problem 2: How do you monitor and fix things?


Not without permission and good reason! There was a second question that I needed to verify, in vSphere's rule book, the Monitor page was to be strictly read only and the Manage page is where you would go to take actions. However I had a concept that was getting a lot of traction, since our scheduling and protecting assets were falling under the manage page, being able to see a problem on the monitor page and then fix it right there, it made sense to me but I needed it to make sense to the administrators. The Canada User group was the first test of this concept and I have had user tests go well, but not this well. For one, being able to integrate into vSphere was a huge benefit because it is one less thing to have run, but the simple idea of finding a problem and then putting the actions right there to resolve the problem was met with a lot of enthusiastic head nods. The feedback was now strong enough to forward this onto to vSphere for approval and when the architect showed off our solution at the VMworld conference, it was met with equally enthusiastic approval.

As a side note, if you look at VMware's new Project Clarity guidelines, the manage / monitor is now gone. It is probably a coincidence, but I did make a lot of noise to get that through. 

Problem 3: Where SHOULD schedules come from?


Where do schedules come from? This was one of the biggest questions we had to answer to get this solution correct. Fortunately this question was just in time to be settled for our yearly visit to Canada to meet with with our users. I was sent to Montreal and Ottawa with two of the architects to present a recap of what was completed in the year past, to look at the current state  of our newest designs and lead a research session designed to probe into the future (and current) concerns of our customers. In total we ran 4 sessions that was attended by 56 total customers. 

This did lead to a crucial design pivot regarding who should create the schedules. It turns out the existing storage admins wanted to retain that ability and only allow the VM admins pick the schedule. This lead to a collaboration effort with our Operations Center team to allow for this change.

Problem 4: Restore?


When the 2016 roadmap was constructed, the ability to restore virtual machines in vSphere was something we wanted to do, but had it planned for 2017. However development was awesome and by the end of 2Q we realized we had the bandwidth to do this and we knew we had customers asking for it. Integrating it into the Monitor panel was easy as it became our hub for problems, we added a view selector to find deleted VM's and created an easy to follow restore wizard.

As an added bonus, the work done here resulted in my first design patent.


Up to Spec

Working with my graphic designer was an amazing privilege, and just not because we shared similar views about the NBA, but it was that even though we worked in different cities and never met for over a year we became friends. Oh, and we played off each others strengths to achieve ridiculous amounts of efficiency. We both worked in Illustrator, and constantly traded UX lessons for graphic design lessons. He left the company in October 2016 for bigger and better things, but I credit him for refining my ability to wireframe in high fidelity. 


Prototyping the user tests

Admittedly this was the most challenging only because IBM's NDA is extremely restrictive on what tools we are allowed to use due to materials being hosted to the cloud... which is almost every means of prototyping these days. That still did not stop me, I was able to create a clickable Powerpoint with hotspots that simulated what Invision does, to my surprise it worked out better than I expected and did generate enough feedback to correct a few gaps we had in the design. 

Once we had a working demo, I was able to run user tests that allowed for more exploration. My favorite moment of the testing was having a user ask a question "Where is feature 'x'?" ...."Oh never mind! I found it, this is so intuitive!" 

Two months after the testing was complete our team was approved for a full Invision license, which will make 2017 easier. 




I certainly had my challenges and roadblocks along the way, this was a legacy team with a lot of legacy habits (some of the code is 20 years old), though I feel I earned a black belt in negotiating with development. But the battle is that as a designer your instincts are screaming "this is all terrible please let me start over!" and this is still a working solution that customers are paying for. So you have to keep the bigger picture in mind and know that these incremental changes are actually making someones job easier. 

It certainly has changed how I feel about design and if you have read this far, then this is relevant. All my life I have been drawn to fixing peoples problems, it is why I got into customer service, then technical service, and even when I was teaching people martial arts. I live for that connection with a complete stranger and they thank you for fixing their problem. IBM does not offer that to me at this time, I am too far removed with conference calls and NDA's, and at best I am talking to someone who is talking about a problem someone else is having.

To design for people I need to interact with them, because then I can tap into my passion.