Speculative fiction design piece on modern dating and AI algorithms.
Problem
The way people connect and date has been influenced by the way technology, specifically dating apps, have been designed. Dating apps are designed to categorize people through filters on height, ethnicity, occupation, and more. However these filters are poor indicators of a sustainable relationship and instigate discrimination by contributing to people’s biases when trying to find matches.
Solution
With this in mind, we knew the most productive thing to do wasn’t creating another dating app, but rather to start a conversation about how we respond to the influence of them. Even though dating apps play a role in the issue, we still have a choice in the way we respond to them.
Insight
Algorithms will always be inherently bias because they are created by humans who are inherently bias. Therefore it is not entirely the fault of the algorithm to blame in people’s decision-making processes. It has more to do with human behavior and individual choices in spite of the algorithim’s influence.
People do not always want to change when presented with their bias. Dating algorithims contribute to physical biases people have when swiping through matches. We tested what would happen if people became aware of these biases and learned to respect the fact that not everyone will want to change.
Dating apps don’t make money finding you a relationship the same way casinos don’t make money when you win. They are designed to keep users hooked.
Brief
This project was for our senior capstone thesis for the Interaction Design program. We had the choice of working individually or in a team to research, design, evaluate, and create an innovation idea to solve a specific problem.
Role
This was a group project in a team of 4. I was the user researcher and ux/ui designer. I designed the survey, competitive analysis, performed outreach and organized participants for user testing, designed wireframes and high fidelity prototypes.
We designed a survey on Google forms to gain quantitative and qualitative data on our target audience. The goal of the survey was to better understand users pain points and their behaviors on dating apps. We sent the survey over Instagram which was shared by SMC’s career services account, Facebook groups, and to friends who have used dating apps.
We asked if people like the swiping concept. We were surprised at how split the answers were.
For our research, two of us attended a speed-dating event hosted by The Zero Date where we rapidly interview a total of 6 people. We wanted to know why singles are seeking alternatives to popular dating apps such as Tinder, Bumble, and Hinge.
Hypothesis
Singled out singles who want relationships are using mixer events to meet people in person instead of using dating apps.
Observational research goals
“When we exist in a culture, and we create what is desirable we have to undo the problematic messaging, we have to keep working on these systemic issues.” - Dr. Jess
Dr. Jess was particularly insightful about the need for honesty and transparency in dating, striving for people to address situations directly and to not fear rejection.
At Config, a conference hosted by Figma, we attended a talk on designing for modern dating by Hinge VP of Design Lindsay Norman. The talk served as a check-point in our research process as we compared and contrasted the pain points identified by Hinge. We shifted our project direction to a more speculative approach after this talk.
The power of the prompt
The Hinge app's design differs from swiping apps by adding friction to the app in the form of prompts. Prompts guide users by tailoring their dating profiles that allows users’ to express their personalities.
User pain points on hinge
Hinge identified four core problems—efficiency, signal (profiles not matching personality), responsiveness, and dissatisfaction with the dating pool. Recognizing these issues as similar to our own discoveries, we wondered whether we could redirect our project in order to go beyond creating another niche dating app.
Gauging chemistry with voice prompts
Hinge introduced a new voice-prompted feature in their app, and the response was positive. The company credits its success to going viral on TikTok, but it also made sure to personalize the experience by reducing the noise (and number of decisions) users had to make.
After weeks of researching the relationship between dating apps and human behavior, we hosted several brainstorm sessions to start figuring our main points. What were some of the most important takeaways for people to know and what did we want to create that could further the conversation on our research?
Affection is shaped by culture
The objects of our affections are shaped by our cultural environments, as well as the affordances of the platforms that facilitate intimate interactions. Dating apps like Tinder play a role in facilitating romance, but it does force users behaviors or actions. There is still a much larger societal component that is more difficult to analyze in modern romance.
Our take on swiping
The swiping concept is great from a design perspective since it emulates the natural sensation of how the human mind works, but it is terrible for mental health. It’s an addictive behavior that rewards instant gratification with dopamine. We thought about different microinteractions that could replace the swiping concept but ultimately decided that our focus should be on creating conversations for others to join us on how we can reimagine something better.
The door for discussion
Our team hosted several brainstorm sessions that were rich in their depth and diversity of ideas. However, we ultimately felt that we needed to move forward with one idea we could use as the door for discussion. Our other ideas were not feasible and would require extensive testing.
Algorithms are biased
All algorithms have biases because they are made by people and people have biases which transfer into algorithms.
The message
People are dependent on algorithms, and they aren’t aware of it. AI predictions show users what their future might be like with potential matches. Will people change their decision-making in dating if they knew their future?
Exposing people to their biases
Exposing how the algorithm categorizes people and showing users their tailored preferences.
The message
Everyone has a type, and the algorithm even knows it. Bias Reminders helps bring awareness to peoples unconscious biases. What if the algorithm called you out on your biases? How would you react to this?
We live in a disposable world
Dating apps and swiping specifically have enabled people to view others as disposable. Swiping is a micro interaction that occurs when you are liking or disliking a potential match. This microinteraction can be executed within seconds which allows for users to swipe a hundred, maybe even a thousand people daily.
The message
People view others as disposable because of the way dating apps are created to function and your money is being exploited to use premium features. Once you swipe left (dislike) on 100 people your trash will become full and you’ll have the option to either review trash or have to pay to empty it. This gives the disliked profiles another chance of getting a match.
Gone without a trace
Ghosting is a direct side effect of the algorithm overwhelming you with thousands of matches. You may unintentionally leave someone hangs just because you have 50 other possible people you’re interested in.
The message
Being more mindful of how we treat others online. When people end conversations with others online without an explanation the other person is often left wondering what happened. This ghosting feature will unmatch you after 2 weeks if the conversation ends and you will have 24 hours to either revive the chat or let it expire.
But don’t worry, you can still view these expired matches in settings under Plan B’s. Cause you know, just in case whoever you're talking to now doesn’t pan out.
We talked to 4 different people who have familiarity with dating apps to better understand their opinion and to start a discussion with them on our concept. The goal of our testing was to validate or rebuke our research assumptions, discuss the 4 features we created, and to get ideas about how to move forward.
Make it easy to understand
The 4 ideas we presented were difficult to understand from an outside perspective. One user told us that we needed to find a way to make our ideas more simple without having to explain how it worked.
Biased reminders may fail
We operated our project under the assumption that dating app algorithms contribute to the bias people share and these biases are problematic. We assumed that when presented with these biases that people would change, however after user testing, we realized that not everyone operates under the lens that biases can be problematic. One of our users even told us that some people may be proud and accepting to admit their biases.
Think about human behavior
After doing user testing, we realized our concepts asked people to do too much. We were creating features without realizing how opposite human behavior is. For example, dating apps are built around swiping because it is instant gratification. Our concepts asked users to go through slow processes of interactions.