UX case study exploratory research about movie and television streaming platform search and display

Previous Next Project

MOVIE & TV SHOW SEARCH

Diving deeper into movie and tv search

Research | Design

My Role

Research/Design

Timeframe

Ongoing

Team

4 students

Tools

Optimal Workshop, Dovetail, Figma, Slack, Zoom

Summary

Finding personally appealing movies and tv shows on streaming platforms can be frustrating. Our team questioned the nature of the frustration. Did it center around algorithms, IA (information architecture), UI (User interface), control, selection, proprietary platform issues, or a mix of all? And which is most critical? We hypthesized that if we could create a more interactive and personal search experience, users would find more of what they like.

normal moments proposed redesign — Statistic from Statista 2022

On YouTube alone, around 500 hours of video content is uploaded every minute.

-Statista 2022

Results

We advanced knowledge of user preferences and expectations for movie and tv streaming search and display.

The Problem

The amount of content to watch is massive and growing. Users have a difficult time deciding what to watch. They are inundated with lists and recommendations, yet feel a lack of personalization. Despite all of the recommendations, our research showed that users still receive the most successful recommendations from friends and family. Streaming platform search and discover remains overwhelming and frustrating.

“...I am always confused what to watch. There are a lot of shows and a lot of recommendations, so it's quite overwhelming.”

Interview participant

How we explored the problem space

Conducted literature reviews and analyzed streaming platform customer reviews

Competitive analysis of streaming companies: Netflix, YouTube, Just Watch, Maimovie, Cinetrii, Trakt/Plex

Talked to users through interviews & contextual inquiries

Received feedback from visual navigation tests

“...Sometimes when I'm looking for a new show to watch, I'll spend my entire time on the treadmill just watching trailers and I won't even pick a show.”

-Interview participant

UX Process

This exploratory research project started by conducting secondary research and competitive analysis followed by contextual inquires. This fundamental information quickly allowed us to pursue different paths of inquiry. Our process created a solid foundation for additional qualitative and quantitative studies in this complex and changing area.

process number one in ux design thinking

Listen

Interview streaming platform users about their experiences.

Learn

Secondary Research

Analyze

Transcribe and discuss findings from interviews.

Create

Create User experience map and visual test questions for 1st and 2nd visual prototype tests.

Test

Contextual inquiry with visual test prototypes

user experience map for a streaming content movie search journey

“Can't remember the last quality bit of content I saw on there. It's all copy and paste documentaries about increasingly uninteresting nonsense....”

-Online reviewer

Questions generated during the research process

What sorts of visual and functional distinctions do platforms make for movies vs. tv shows?

Dovetail tag board for video search project

How and in what ways will AI change streaming platforms and movie/video search?

How much loyalty do users have towards a particular streaming platform?

Would algorithmic transparency provide a better search experience?

What might radically different search-result-alternatives look like?

How can search be more interactive, engaging and personal?

“I find that the [streaming platforms'] categories don't exactly fit my taste. I want to be able to choose one and search deeper from there..."

-Contextual Inquiry Participant

Design + Test

In addition to conducting interviews and contextual inquires, we designed three iterations of visual navigation tests to gather user feedback. We started with 11 concepts for the first test and narrowed it down to 6 for the second test. The third test we developed a hi-fidelity, interactive version of 1 concept.

Usability test for streaming content medium fidelity

Usability test 1

Administered through HTML page
Moderated
10 participants
Low-Fidelity
static

Usability test 2

Administered through Figma link
Unmoderated
9 participants
Mid-Fidelity
Interactive

“I feel #11, 3d bubble, would give me curiosity to explore"

-Contextual Inquiry Participant

Insights

Most participants did not want a movie/tv search process that made them think too much.
Users could be divided into two groups "Seekers" and "Statics" .
Making features or systems similar to ones already used was important to over half of users.
Users need help remembering and storing movie info so it is readily available when ready to watch.
Watching a feature length movie is an investment of time and users want to feel good about the investment

“...In my recommendations there are a lot of cringy movies and series. I feel like I get that a lot.”

-Interview Participant

Slider portion of test

Usability text for streaming content medium fidelity — Sliders on 2nd navigation test

Slider test 1

Administered through HTML page
Moderated
10 participants
Low-Fidelity
static

Slider test 2

Administered through Figma link
Unoderated
9 participants
Mid-Fidelity
Interactive

Slider test 3

Not administered yet
Moderated
10 participants
Hi-Fidelity
Interactive

“I want to, for example, search by director and then see all the connections and associations to and from that director...”

-Interview Participant

Constraints

Limited time and budget
Difficult getting enough participants for contextual inquiry
Too wide of a scope
Exploratory in nature with no conclusions made

Reflections

This exploratory research project generated great ideas and feedback. However, nothing was definitely solve and more studies are required to gain additional qualitative and quantitative data. Using a larger sample size with a diverse group would also be helpful to achieve more generalizable results. One important learning is that conducting an unmoderated interactive test, as the 2nd test was, could have been much more effective and data-generating if moderated. Though it was accompanied by a google survey, the test was likely too complicated and intricate for an unmoderated test.