Photo of the 2016 model of the Apple TV device (right) with the Apple TV remote (left). While beautifully simple, the device suffers from some heuristic errors. The remote is the subject of our research.
Brand: Apple
Project Goal: Find a design tool to address Apple TV remote issue.
Summary: The exploration of Apple TV’s remote design and strategies on how to address issues of usability.
Team: I was the team lead of two others on this group project for Psychology of HCI project
Responsibility: User Research, Testing, Persona Development, Interaction Design, Prototyping, Visual Design, UX Design, 3D Printing
Tools: Sketchbook & Pencil, Google Docs, Photoshop, Maya
Apple TV (2016) Remote UX Review & Application of Asymmetry in Design
Introduction
My goal for this Human-Computer Interaction group lab assignment is to explore and improve the user experience of the 2016 revision of the Apple TV remote. I have a particular disdain for this particular remote and the team as a whole decided to explore the issue further. While Apple is often praised for its industrial design and innovation in its product lineup, it often comes at some cost to the users experience. Rather than focusing on heuristics, Apple chooses to go more for what “looks good”. Below is our exploration into how one Apple product could improve with a few simple changes in its industrial design.
Issues with current Apple TV remote design
The symmetrical design of the Apple TV remote makes it difficult for the user to distinguish whether they are holding the remote correctly.
There is little to indicate the correct way to hold it without looking at the device, or testing the button, which may lead to potentially problematic results. Since each button serves a different function, they cause an effect that is dramatically different than what the user may have intended.
When holding the remote in the wrong orientation, the user’s palm rests on the touchpad, which can also cause unintentional inputs.
Initial UX Testing
The user base for the Apple TV range from children to elderly, with a majority of users being in their 20’s-30’s.
Remote is intuitive to use on the interface it was designed for.
Point of failure at first point of contact, leading to 50% chance of needing to make a corrective action.
Touch-pad is intuitive, however sensitivity varies, and can not be used if finger is damp.
Siri tends to fail more often than not, leading user to try and try again.
There is no streamlined way of finding the specific piece of media you are looking for.
Design Goal
My goal is to design an Apple TV remote that limits user error when picking up the remote without looking at it; the “dark room couch potato” scenario, if you will. We feel this can be done by simply adjusting the symmetry of the remote, but we will not limit our research to just this aspect in our design.
Photo of a page from my sketchbook with some concepts on the left; figuring out the shape that would address the issues found in the Apple TV (2016) Remote. On the right are two high fidelity drafting of the 2016 remote next to an alternate design that will be 3D printed.
Design & Testing
Competitor UX
There is no “perfect remote” that currently exists, and as a result, there is no gold standard to follow. There are features in other remotes that feel good, such as a subtle notch taken out under the primary navigation, or design with the center of gravity being taken into account. Overall, remotes designed by competitors are clunky, and filled with buttons that appear to have redundant features because of the lack of unity between all possible television peripherals.
Apple TV UX Review
The list below mentions where the Apple TV remote gets things correct in terms of UX, and how they could be improved.
Remote is intuitive to use on the interface it was designed for.
Point of failure at first point of contact, leading to 50% chance of needing to make a corrective action.
Touch-pad is intuitive, however sensitivity varies, and can not be used if the finger is damp; for example, after holding a glass with condensation.
Siri tends to fail more often than not, leading users to try and try again. Seems to improve with time, but can be frustrating.
There is no streamlined way of finding the specific piece of media you are looking for on any of the available streaming platforms.
Design
The steps I took to improve the design of the 2016 Apple TV remote (as seen in the sketches above) was changing the shape of the top and bottom in a simplified and more comfortable approach to the remote. The main issue we had with the 2016 Apple TV remote was it’s symmetry, leading to difficulty when telling the difference between the top and bottom of the remote. In our initial redesign this resulted in removing the touchpad and instead adding a click wheel to the front, changing the flat bottom to a cylinder wedge, as well as moving the buttons from the middle of the remote closer to the front. We decided to also remove the home button because it seemed redundant as the remote already has a ‘Menu’ button that can take you back to the home screen. The reason we chose to stay away from a flat bottom to the remote was to avoid difficulty picking it up off of flat surfaces. When deciding to make the top thicker than the bottom part to help with picking it up, it also proved to be more comfortable in the palm of your hand.
Design Features
A.) Change the shape of the remote to ease directional recognition
B.) Made asymmetrical to distinguish the front from the back
C.) Remove the touch surface for a click wheel to help users with fine control issues
D.) Make it easier to pick up the remote from a flat surface
E.) Improve hand feel
F.) Simplify buttons to remove redundant features (Home button)
Testing Plan
The tests we performed had a goal of measuring the improvements of our initial redesign. The number of people we were able to test was limited and ended up getting results from eight different people of various ages. We asked our participants to perform a blindfolded test: each remote will be placed in front of a person in ten predefined orientations, after which they will be asked to pick up the remote, correctly orient it and press the ‘menu’ button. We then compared the rate of user error in time spent trying to correctly orient the remote and then click the ‘menu’ button. Along with that, we had them fill out a Likert scale (as seen below) to measure the ease of navigation, ease of handling, as well as preference between the click wheel and touchpad.
Testing
Test ergonomics - Compare rate of user error when blindfolded
Ease of navigation comparing click wheel versus touchpad (Likert scale)
Compare ease of handling and comfort (Likert scale)
Photo of original Apple TV (2016) remote along with a 3D printed analogue along with test design print. The two 3D prints will be used for user testing in hope of testing the theory of an asymmetrical design reducing user error when using the remote.
Conclusions
The key takeaway from the test is that the asymmetric design of the redesigned remote helped our testers use the device. This is shown in both the timed blindfold tests, where testers were more likely to breach the 5 second time limit at a total of 20 times of the total 80 tests, vs the redesign, where it was less likely to go over the limit at only 5 times of the total 80. This was also recorded in the short answers, where users described their preference for the redesigned remote. Lastly, they confirmed that asymmetry improved device orientation recognition, and even helped the testers pick up the device from a flat surface.
Time to find the ‘menu’ button on the 2016 Apple TV remote
>5 second limit breached 20 times. (results below)
Time to find the ‘ menu’ button on redesigned remote
>5 second limit breached 5 times (results below)
Breaking this data down further, the average time a tester took to find the menu button on the existing Apple TV remote was approximately 36.7 seconds versus the approximately 28.2 seconds it took them to find it on the redesigned remote. That is a delta of approximately 8.5 seconds to identify the shape of the remote and to find the menu button. The largest delta was found when the remote was placed upside down for the tester, averaging a approximately 12 second faster.
Although the redesigned remote successfully indicated proper orientation, it was not the only aspect we aimed to improve. The removal of the touch-pad in favor of the click-wheel was another change we made to the remote. However the results of this change were inconclusive as seen below:
One of the comments that we received during the test stated that due to the remote being very thin on one side it was harder to grip and pick up from that end. While it was still an improvement from the current (2016+) design it was a valid point that we took note of.
Another aspect that turned out to be inconclusive was the simplification of buttons. We removed the redundant ‘Apple TV’ button, and moved the less frequently used Siri button lower and made it smaller. Judging by the results and by some of the interactions we had with the testers, not all of them seemed to notice this change. The results can be seen in the graph below.
The redesigned remote achieved the main goal of improving the recognition of the front from the back of the remote and cut down the time it takes to do so. (see test results in figure below)
Perhaps the most remarkable finding was how much the asymmetrical design improved the testers' recognition of the orientation of the remote in their hand, just by feel alone. Nearly 90% of participants strongly agreed that the asymmetrical design improved orientation recognition with one only one agreeing that it does. This was our primary goal with our redesign, and encouraging for our future revisions.
Full Circle
My key takeaway from my first test was the importance of having an asymmetrical design. As I found in my survey, an overwhelming majority of my test subjects found that having some asymmetry in the design helped with recognizing the front of the device from the back. This is especially important for a handheld device, such as a remote. We could especially see this design implantation helping people who are hard of sight, or blind, interact with the remote. So yes, overall, I reached my primary goal of creating a remote that could quickly be picked up and pointed at the TV without looking at it.
Photo of what a revision two of the Apple TV remote would look like. The primary changes would be shifting the weight of the device to the back to make it sit in the hand a little better, bring back the touch-pad to see if that changes shape recognition, and to add a large bevel on all side faces to help the user pick up the device.
The Path Forward
What’s Next?
After a successful first test, I learned how important asymmetry was to a device handling and recognition. Building off of what I had learned, I wanted to create a second revision of the Apple TV remote, addressing some of the feedback I had received. The design features changed from version 1.0 to version 2.0 of the Apple TV remote redesign are as follows:
Added “pistol grip” to the remote to help with grip and shape recognition.
Fat end of the remote reversed to fit in the user's palm better.
Click-wheel replaced with touch-pad to remove confounding variables.
Thinned area below touch-pad to ease button presses.
Rounded portions of the top of the remote to facilitate “rocking,” easing edge cases where the remote is placed upside down on a flat surface.
Changing the center of gravity of the remote to fall back into the user’s hand rather than forward out of a user’s hand.
Final Thoughts
A device can be both intuitive to use but still fail to be recognizable.
A symmetrical design relies on recall rather than recognition.
An asymmetrical design is a good way to promote recognition.