Artist on X
Kelly McCraw and I conducted a study on the social media platform "X" with a focus on making it more accommodating for artists. We carried out user research to understand artists’ pain points and, using design thinking methods, came up with possible solutions to their concerns.
5 weeks
Team of 2
Role:
UX Designer
My Roles:
- Recruiting Particpants
- Competitive Research
- Designing & Creating Wireframes
- Designing & Creating the Mockup
Shared Roles:
- Writing the Survey
- Running Usability Tests
- Problem-sovling
- Decision Making
-
Survey​
-
Affinity Map​
-
Competitive Research
-
Personas
-
Empathy Map​
-
Insights
-
Wireframe
-
Usability Test
-
Mockup
-
What did I learn?
-
Measurement of success?
-
What's next?
-
Thanks
The Problem
The social media platform X has been losing active users in the past year. We believed focusing on a community within X's users and addressing their concerns could reduce platform abandonment. However, as we don't work with X, this case study is entirely conceptual.
​
We decided to focus on a community that we are familiar with: Artists. This community includes visual artists of varying skill levels, commitment levels, and follower counts. Some treat their role as an artist as a casual hobby they do on the side, while others are professionals who make their living off the art they create. Despite their differences, they all share one thing - they post their art on X.
How can we change the social media platform "X" to be
more accommodating for Artists?
The Plan
Methods
Throughout this case study, we used methods like user research, surveys, competitive research personas, empathy maps, usability tests, wireframes, and prototypes.
Approach
We started with user research to understand our users' needs. With this knowledge, McCraw created personas and an empathy map to define our users and their problems. Using the insights, I designed a prototype for usability tests. I refined the prototype based on feedback and created a realistic mockup.
Process
Discover
Define
Ideate
Prototype
Test
Understand
Survey
Our first step was a survey. We recruited 32 participants to answer 15 questions on Typeform. Our primary goal was to find the pain points Artists have with X. We also wanted to learn about Artists’ general opinions of X and other social media platforms.
As an Artist, what on X do they find unfriendly?
What’s their favorite social media platform?
What do they like about the platforms they use?
Interesting findings from the survey:
"Do you think X/Twitter is art-friendly?"
Reasons artists stay on X:
-
Easy to interact with their audience
-
Can turn off the algorithm
-
The people
​
​
Reasons artists post their art online:
-
Sharing
-
Expression
-
Recognition
-
To attract work
-
Maintain an audience
Affinity Map
We created an affinity map with the information from the survey with physical sticky notes. The notes are organized into groups of different aspects of an Artist’s experience on X.
The notes from “Social” and “Motives” are the reasons Artists spend time on X, “How Art is Shared & Viewed” notes are dependent on the mechanics and features of X, and “Threats” refer to Artists’ looming concerns.
Competitive Research
I picked out the social media platforms people mentioned in the survey and compared them. I tried to give a gist of each platform while also covering themes expressed in the survey. Those themes include flexibility and options in posting, algorithms, and threats of art being used to train AI images. Please note that all pictures online can be used to train AI image generators, but Instagram and DeviantArt specifically use images on their websites to train their own AI generators.
Of this list, Instagram seems to be X’s biggest competitor. However, despite its large monthly active user count and being an image-focused platform, Instagram goes against a lot of Artists’ values. Instagram has no alternative to its algorithm-based feed, and users’ public images on Instagram are being fed into Meta AI, an AI image generator. There are also restrictions when it comes to posting, such as the inability to post from a desktop and making scheduling posts a premium feature.
Instagram was not popular in our survey. Three different participants wrote, “It’s not Instagram,” when asked about the positives of X. However, it’s important to relay that we got our participants through X. I’m sure we would’ve heard a different side if we recruited users through Instagram.
Competitive Research
I picked out the social media platforms people mentioned in the survey and compared them. I tried to give a gist of each platform while also covering themes expressed in the survey. Those themes include flexibility and options in posting, algorithms, and threats of art being used to train AI images. Please note that all pictures online can be used to train AI image generators, but Instagram and DeviantArt specifically use images on their websites to train their own AI generators.
Of this list, Instagram seems to be X’s biggest competitor. However, despite its large monthly active user count and being an image-focused platform, Instagram goes against a lot of Artists’ values. Instagram has no alternative to its algorithm-based feed, and users’ public images on Instagram are being fed into Meta AI, an AI image generator. There are also restrictions when it comes to posting, such as the inability to post from a desktop and making scheduling posts a premium feature.
Instagram was not popular in our survey. Three different participants wrote, “It’s not Instagram,” when asked about the positives of X. However, it’s important to relay that we got our participants through X. I’m sure we would’ve heard a different side if we recruited users through Instagram.
Interpret
Personas
We categorized our survey participants into The Influencer, The Professional, and The Hobbyist. The Hobbyist pain points weren’t as defined as the others, so we stuck with the first two. The personas are based on our survey and real artists on X.
Having the personas helped pull together and organize the variety of information we got from the affinity map into memorable people whose actions and opinions make sense.
Empathy Map
McCraw made an empathy map for specifically Ian the Influencer.
The empathy map made artists’s grievances more tangible. It wasn’t until reviewing the personas and empathy map together that I fully understood the hierarchy of the pain points.
Insights
Through the personas and empathy map, it became clearer what artists on X cared about. We narrowed it down to the top 3 values.
AI Art
Algorithm
Presentation
X's Algorithm
X's algorithm is a subject that passes my knowledge and reach, but I'd be happy to gather the connected parties to relay the feedback we've received. In this meeting, I would share that artists on X feel X's algorithm:
-
Kills posts after two days
-
Favors what's trendy
-
Require the artist to post during specific time zones
-
Does not work with hashtags appropriately
-
Ignores small accounts
AI Art
Artists on X are vocal about AI art due to the art theft required to produce AI images. X has already surpassed other social media platforms, such as Instagram and Deviantart, which are complacent with images on their websites to be used for training AI generators. If X wanted to take a firm stance against AI art, it could include tools like Glaze and Nightshade among their filters. These tools protect users' art from being trained by AI art tools.
If I were part of X's team, I would ensure that artists' feelings towards AI art were conveyed to the appropriate shareholders and initiate discussions on how to address them, making sure that the voices of artists were heard and respected.
Presentation
In the survey, the participants commented on how their art was presented throughout the website, like in their posts, profiles, and media tabs. Ultimately, the artists wanted more control over how their art was seen. In the next section, “Construct,” I design a feature that gives artists control over how they present their art in their posts.
Construct
Currently, on X, the image to the right is what posting four images looks like on mobile. You can change the order of the images by sliding them, but you don’t know what the layout will look like in the published post, let alone have any control over it.
​
I set out to create a version of the posting page where the user can see how their images will look published and control the layout of the images.
Through the competitive research, I knew Tumblr (on the right) was very flexible with how users presented their images. Users just had to drag the image to change the layout. Tumblr’s method was my inspiration when designing a similar feature for X.
Usability Test
For the sake of time and flexibility, I went with a low-fidelity paper wireframe for our usability test. I drew the phone on a sheet of paper and cut out different rectangles for the images. Since the images can be various sizes between the layouts, I cut out several sizes for each. I created a stack of cards with different layouts, and the idea was to have the participant match the cards by moving the images in a similar way that images are movable on Tumblr. If the user drags an image to the side of another image, the second image should move to make space for the first image. If an image is dropped on top of another image, those images should swap places.
The goal of the usability test was to gauge the intuitiveness of the feature.
Could the participants recreate each layout without aid?
The answer: For the most part, yes. Overall, the 6 participants understood how the layout would change once they moved an image. At times, the participants were inefficient and had to make several moves to reach their goal, but they ultimately reached their goal.
On the flip side, the participants did struggle at times. The idea of dropping an image right under a single image to push the dropped image right under the other image instead of under all images (what's going on in the left image) was not intuitive, nor was it easy to explain.
This became a problem during the last layout (shown below), where the participants had to change their four-tall-images square layout to a four-squat-images square layout. Some tried double-tapping, pinching, or deleting images, while some couldn’t think of what to try. The answer is to drop the bottom images on the bottom edge of the image above it. This clearly was not intuitive.
What I took from the usability test:
-
Having two different ways to have four images make a square was too confusing. It’s best to cut out the four-tall-images layout.
-
The participants didn’t always know what they could do. Perhaps adding visual feedback would solve this.
Mock Up
I took what I learned from the usability test to create this animated mockup in Photoshop. It shows different layouts, how to create them, and the UI around the feature. Changes were made since the wireframe. Tall images in a 4-image square were no longer possible. Visual aids in the form of blue lines were added to show where images could be dropped.
​
I showed this mockup to participants from the usability test. The participants weren't particularly wowed, but they saw the improvements.
​
"The blue lines help make it make sense."
​
"I get it now, seeing the C be placed under the D. Seeing it like that helps."
​
"The presentation is really nice."
​
​
​
Reflect
What did I learn?
It’s not a new realization, but it is always surprising to see unexpected survey responses. It reminds me that I can’t predict what people think, and, to extend that, there’s a lot I don’t know. It’s my instinct to plan before I gauge the target audience, but I need to work on slowing down and trusting the process. Otherwise, I might get disappointed when users disagree.
​
I wasn’t disappointed this time because enough artists cared about how their art was presented to warrant me making a prototype, but I didn’t expect artists to care more about the algorithm and AI.
Measurement of success
Despite this being a conceptual project, we were able to get positive, qualitative feedback from the usability test participants on the mockup. That shows that we listened to our users and were able to address problems. If this were live, I would measure success by how many users use the feature and how often it’s used. The rate of people leaving the platform is too complicated to reflect this feature purely, but I would keep an eye on it.
What's next?
If I were to continue this project, I would create usability tests and mockups for alternative ways to offer flexibility in how art is presented on X. For instance, I would consider adding options to the media tab for different viewing modes. If this weren’t a conceptual project, I would be very interested in what shareholders say about artists' complaints about AI art and the algorithm. I would love to have those conversations with the artists’ best interests in mind.
Thanks
I am very thankful for my partner in this project, Kelly McCraw. Her personas and empathy map put the artist’s pain points in perspective for me. It made me realize that art theft and the algorithm were more significant concerns than I initially thought. I am also very thankful for all our participants from the survey and usability test. This project was only possible because of their aid, and I loved reading/hearing their opinions and feedback.
If you yearn for research,
Comfort Games is the
place to go!