User Experience Research
of
YouTube
YouTube is the most popular video sharing platform in the world. With billions of videos across millions of channels. With such a huge selection, how well do users find what they are looking for, navigate the platform and use its UI controls.
UX researcher
survey designer
script writer
interviewer
Keynote, Google Forms, Google Sheets, Adobe Illustrator, iPhone video recording
independent
To understand how to keep improving the YouTube platform and to know what issues and features to prioritize for stakeholders, I sought to answer:
how people currently use the site?
what issues people have using the site?
Using a Lean approach, I had to act fast to—
First I created a survey according to good design principles, such as giving users an 'other' option to write-in. Then I conducted an online survey over a few days to gain an overview of who the users are, their perceptions, feelings, and how they use the platform.
When I launched the UX survey in Google Forms, my goal was 15-20 participants, but I wasn’t certain if I could get that many in a few days. However, by asking participants for referrals, I exceeded my goal. Aside from most participants having four year degrees, I was particularly happy to get a diverse group.
86% of participants used search to find what they are looking for
73% of participants have shared a video directly from the site
32% of participants' biggest frustration was specific to the individual
The quick survey worked as expected to give me an overview of who the users are and how they use the site. It also showed me that prominent features of the platform are overlooked, so I decided to further investigate that during usability testing. Therefore, I refined the questions to learn general use and specifics.
How people get started with the site?
How people use the site?
What controls people use on the video player?
How people use Youtube Channels?
What issues people have using the site?
Because the research goal included learning about both user behavior and thinking, I chose to conduct a usability study with think-aloud.
NDAs would not be needed because what we are testing is public. However, parental/guardian permission would be necessary for anyone who was not a legal adult.
My goal was to recruit five current YouTube users. The primary plan, which was successful, was to recruit participants from those who completed the quick survey. My backup plan was to reach out to people by email, social media, and word of mouth.
Thankfully, based on their initial survey answers, demographics, and willingness, the initial survey also served as a screener. In that way, participants in scope of the research study were found.
Then I made time within a week to conduct the sessions.
So that other researchers could replicate the study in the future, I planned to document my analysis process as I executed it. To maximize relevant data collection, ways were considered to make sure the data would be reliable, how to get insights from the data, and who the findings would be shared with.
Based on the research goals, five scenarios were devised so that users could perform the tasks. The expected 'correct' answers for each task were noted.
The script was written to make sure all the steps were covered, non-leading wording was used, and so that tasks were in the same order for each participant.
There was an opportunity to quantify each participant's feelings by surveying users directly after usability testing, so I created another survey.
The study was done in each participant's own environment to have their experience as typical as possible. Each session lasted about twenty-five minutes. Following the script, I asked permission to video record which made made notetaking less stressful and made the testing much easier to moderate.
After my introduction, before jumping straight to tasks, I eased the participant in by asking some open-ended questions like:
What on the site is most important to you?
What are your initial thoughts about the screen?
What captures your attention?
Then we stepped through five scenarios. The tasks were grouped into scenarios both for a more natural progression and to provide little breaks to debrief the participant on what they just did. During debrief, I asked more scripted questions to gain direct insight into their knowledge, thoughts or feelings about the tasks.
At the end of the session, I asked the participants to complete a satisfaction survey so that I could quantify their thoughts and feelings about the tasks.
4 of 5 participants do not use channel pages and 2 of 5 did not know what channel they were on.
Liz — "My biggest issue is when I can't find a video again."
3 of 5 participants do not use YouTube's mix/playlist even though it takes up a sizeable amount of the screen area.
Tess – "There are so many unintersting videos."
4 of 5 participants used the time timeline control in the video player. However, only 2 of 5 participants knew how to rewind besides using the timeline.
Nancy – "I don't know about all of them, I only use pause, volume and full screen."
display the channel label more prominently so that people can find them, and find more of what they are looking for.
try including 'channel' as a filter in the mix/playlist.
research why mix/playlist is not used.
consider redesigning the mix list and/or playlist, possibly combining the two.
make timeline easier to control, including for rewind.
I would check that the site and video player is accessible to people using a screenreader to navigate the site.
By using Toptal's color blindness simulator, I would check that the site including content is accessible for people with various types of color blindness or color vision deficiency.
With stakeholder input, I would plan monthly iterative testing so that design and development teams learn that research is ongoing and so that same-day decisions can be made on which issues will be fixed before the next month's testing.
Based this research, an Eye tracking session would help determine if people even look at the channel label, playlist, and the video player buttons and controls. We would answer if it’s a lack of seeing or of understanding ‘Why most users don't use these features?’.
Depending on relevant stakeholder needs, I would pitch, post, and/or make a presentation to share this UX research.
It was relatively easy to keep emotions in check with adults, but with a younger person I found that more difficult when they got excited. Next time I would be more aware to keep my own reactions in check.
If possible, I would have parents wait in a separate room so they don’t influence the study results of adolescent participants.
Although participants may say they know how to use a product, usability testing is a great way to learn how well they use the product, and how familiar they are with its features. And thus, UX research can also uncover gaps, challenges, and opportunties for the business.
