Peter Ponomarev

Background

Windows 8 was a major departure from previous versions of Windows because it introduced a touch-friendly core experience that scaled from small tablets to desktop PCs. Though this was mostly an addition onto the familiar desktop experience, it was felt imperative to add features that would delight devoted Windows users, who are classified as “enthusiasts”.

Task Manager was seen as a key investment area, as it is considered an enthusiast feature and is used by more than 500 million people worldwide.

INCEPTION

Process phases explained

The first question to ask was “Who are enthusiasts?” I partnered with Product Planning and User Research to better understand this user segment. Product planning had done some initial audience research and developed some rough themes, but the themes weren’t yet actionable from a UX standpoint.

I partnered with User Research in phone interviews and field visits with users who met the basic criteria of an enthusiast. We merged our findings with the earlier work by Product Planning. While there was a wide variance in the tastes, preferences and abilities of study participants, we were able to distill our thinking into four key motivations that all enthusiast shared:

These principles were used in ideation sessions where we generated idea cards for enthusiast features. Each idea card was grounded in the enthusiast motivations and used a common template. One of my idea cards was for modernizing the design of Task Manager:

ELABORATION

Not surprisingly, I was assigned to Task Manager. One of the first things I did was to analyze the existing application, both through lab studies of current users and analysis rooted in our UX principles.

In addition to the issues we noticed in the current application, there were related features that were missing. One was managing which programs start with Windows, which was buried elsewhere in the system. Another was viewing application resource usage over time, which is especially important in metered network situations, e.g. a tablet with a 5G connection.

Next, i began sketching potential new ways to modernize, reorganize and combine the functionality of Task Manager. I started with pen and paper sketches to work through basic concepts, then progressed to rough wireframes in PowerPoint, which made for effective collaboration.

After these explorations the conceptual outlines of the interface were becoming clear. There would be a simple and an advanced view. The advanced view would be a tabbed interface with visual indication of process behavior. The advanced view would have a Performance screen with unique graphs for each core resource. The task now was to dig into data visualization, wayfinding and visual signatures.

My first focus was the Processes tab, the functional heart of the application. The goal was to visually indicate which processes were problematic without the need for reading words or numbers.

From a usability standpoint, we wanted to validate that the heat map was effective in quickly guiding users to the correct programs to close. To test this we conducted a lab study utilizing an eye tracking camera.

I was simultaneously looking into the visual identity of the graphs in the Performance tab, while also improving the information architecture, wayfinding and visual balance. Windows 8 was embracing the Metro design aesthetic, so I wanted to look for ways to make the information more beautiful.

While doing lateral research into possibilities, I came across the work of Nicholas Felton. He publishes a yearly “Feltron” yearbook of this life. His graphics were inspirational, as they showed an elegant sense of typography, balance through asymmetry and minimalism.

This approach seemed to largely align with our goals, but it needed to be tested, as there was an implicit assumption that enthusiasts were primarily used to dark, Battleship-type interfaces. So, I constructed a “visceral response” lab test with 30 examples of information graphics, ranging from Windows 95 control panels to the Feltron example shown above. Participants were given 5 seconds to rate each example from 1 to 5 — we deliberately did not allow time for analysis, opinions or commentary. After running the test with numerous participants, the Feltron design (above) won overwhelmingly. This gave us the confidence to move forward with a similar approach.

The first versions of the Performance tab used only one graph color for all of the resources.

The Performance tab was going in the right direction, but it still didn’t feel right. Each resource didn’t feel unique because all the graphs had the same color. The switcher area had too much weight. While the switcher was visually secondary (2nd column), functionally most such controls are found on the left. After many iterations I settled into the final set of designs. The switcher is smaller and on the left. Each resource has its own color and unique footprint, but all resources share the same asymmetrical layout template:

The two showpieces of the new Task Manager were now in place. With those pillars in place the team went about fleshing out the remaining tabs and states of the system. This included:

• Replacing image names like “iexplore.exe” with friendly application names like “Internet Explorer”

• Reducing the initial simple process view to its bare minimum to increase user success. I removed all table heading, rule lines, tabs, readouts, etc. Only 3 things remain. A list of a user’s running apps, the end task button and a “more details” control.

• Moving the control of startup items from Configuration Manager to Task Manager

• Adding an App History tab showing resource usage over time. We repurposed the heat map seen in the Processes tab.

• Re-working the Users tab, which is basically the Process tab, but broken down by active users on the machine

• Visual re-touches to the services tab

• Maintaining the legacy Processes tab, but renaming it “Details”. This was necessary to have full back compatibility, but it was de-emphasized in tab order.

Here is the final set of designs prior to the construction phase.

With all of the screens and tabs designed, it was time to redline everything for construction.

CONSTRUCTION

Construction was moving very smoothly, as much of the core functionality had been constructed in a prior milestone. This freed up Design and Program Management to strategically enhance the product in very targeted ways.

All of the enhancements were focused on the performance tab:

1. Animate the gridlines in the Performance tab. I devised a formula to determine the minimum number of vertical lines based on window width so that the graph would appear to be moving.

2. Enhance the Networking screen by showing a change in bandwidth over time. I invented a concept I called “The water line” which would animate deliberately to show the user that bandwidth was changing over time.

3. Modify the CPU performance screen when there are more than 64 logical processors. Instead of subdividing the line graph, past a certain threshold it switches to a heat map based on the CPU blue color.

DELIVERY

As a final quality check, we mapped all of the new features in Task Manager against the four enthusiast motivations we had identified in the inception phase. These were the touch points that we hoped to see reflected in public feedback.

Once Task Manager had been made available through developer previews and beta releases, it was time to market the feature to the world. Program managers and the president of Windows wrote articles covering the new features. Once the beta releases and onward were available to the public, the feedback about Task Manager poured in and was overwhelmingly positive. For my efforts I was awarded 8 patents.