Product Design: Dating App Soul and its Matchmaking Algorithm
Explore the Chinese online dating market as we dive deep into Soul, China’s AI-driven dating app, and assess its strengths and weaknesses.
In the online dating industry, a product must be effectively designed to offer a great user experience and attract more traffic and users. This article will analyze the competitive landscape of online dating apps in China and summarize the key development direction of Soul’s matchmaking algorithm.
The Purpose of Competitor Analysis
After setting up an online dating profile on any platform, we will first be matched with other profiles. This is followed by mutual interaction and ice-breaking activities, which will eventually lead to an effective relationship. While these three stages are crucial to precipitate online social relationships, the matchmaking process sets the foundation for relationship building between strangers. A good match can enhance user experience and allow the following steps to take place quickly.
Therefore, this article will dive deep into the matchmaking process of Soul and its competitors. The competitor analysis will allow us to map out the strengths and weaknesses of Soul’s matchmaking process, identify critical areas for improvement and suggest related prototypes.
Product Description
1. Overview
2. Product Flow Diagram
3. User Analysis
a) User Attributes
Soul advocates in-depth social interactions, a mode preferred by female users. In terms of demographics, the ratio of male to female Soul users is close to 1:1, an ideal ratio that can enhance user experience. The platform’s target user is Gen Z, as its largest user group is 24 and below. In fact, 63% of its users are under age 35. As a result of the platform’s youthful demographic, most of its users have a high school or lower qualification.
Geographically, around 60% of Soul users are from first and second-tier cities. More than 70% of the users are relatively well-off, and 35% are in the middle- and high-income groups with a strong purchasing power.
b) User Needs Analysis
According to the Gen Z Social Report by Soul, about half of young people use social media to find people with similar interests and browse interesting content. To them, social media is not limited to interpersonal interactions. Social media also provides a platform for young users to consume content, expand networks and relieve stress.
The study showed that Gen Z’s social choices had defied conventional wisdom in terms of user psychology. More than 70% of users believe that similar interests and values are vital to building relationships online. Only 10% of users say that physical appearance is an important criterion. When it comes to user behaviour, it was found that young people are inclined to record moments of their lives and showcase their talents on social platforms.
Gen Z’s user needs can be summarized into four categories:
- Dating and making friends: This is the most common reason for strangers to interact. The Gen Z Social Report shows that nearly half of the users use social apps hoping to fall in love or simply finding people with the same interests. Soul is a platform that focuses on in-depth interactions with various matchmaking methods and a customized feed to better meet the users’ needs of developing personalized and high-quality friendships.
- Sharing: Soul’s plaza is designed to enable Gen Z users to share content anonymously. This allows users to be more open about sharing their thoughts.
- Self-expression and networking: Gen Z expresses themselves by posting life updates, sharing their views and showcasing their talents. As this is a two-way interaction, Gen Z users also understand other people’s views and meet new people on the platform, expanding their network.
- Entertainment: Gen Z has a lot of free time, and they participate in many forms of leisure activity. Soul caters to such needs by providing many interactive activities for their target users. The plaza feed, matchmaking function, and group chat parties engage Gen Z in many ways.
Competitor Overview
Soul
1. Matchmaking Function Analysis
a) One-way selection in “Planet”
The “planet” in Soul’s home page will rotate to show the Soulers who are currently online. By tapping on the blue dot, you can check out a user’s profile page. There is a filter function in the upper right corner, which allows you to filter the users’ gender.
b) One-on-one matchmaking
c) Profile Tags
Soul introduced the profile tag function to help users receive more accurate matches.
Profile tags are set in homepage settings. Categories of profile tags include interests, occupation, unique experience, characteristics, favourite music, movies, games, books, animation and sports.
d) Meet Me
This feature matches users who tap on the “Meet Me” button. After waiting to be matched, users will see a chat page that shows the matching percentage of both users. This percentage is based on the similarity of both users’ profile tags.
e) Audio Call & Soul Cam
Soul’s audio call matchmaking has a limit of four minutes. Users can pay 10 Soul coins to add ten minutes to the call. Before the call runs out of time, users can choose to reveal their profile, allowing the call to continue without any further limitations. Otherwise, the call will naturally end in four minutes.
The Soul Cam is a video call function similar to an audio call. Users can set video filters to mask their faces before the call starts. Users can smile at the front camera during the call or tap on the like button to light up the “like” buttons. If all buttons are lighted up within 4 minutes, the users can continue the call and request the other user to remove the filter that covers their faces.
f) Love Hotline
The only way users can use the love hotline function is by turning it on and waiting for the system to match someone for them. When the love hotline is on, users will be notified when there is a match nearby. The screen also shows the distance and direction of the matched user. Both parties can choose to continue or reject further interactions.
g) Group Chat
In Soul, users can create or join group chat rooms. Each room will have a “popularity challenge”. When the popularity value of the room reaches 600, the challenge will be considered successful, and the room will be publicized and displayed on the Soul Power list on the home page.
2. Summary
Soul introduces various matchmaking methods that are not focused on physical appearance (Meet Me, Audio Call and Soul Cam). This fulfils the needs of users to interact in different scenarios.
However, the current matchmaking process remains inaccurate. The matching percentage based on algorithm and profile tags does result in better connections between users. As the matchmaking process mainly relies on algorithms, users lack the autonomy to filter matching results on their own.
In addition, the matchmaking process is also not highly monetized. While there are paid functions such as the “same city card” and “speed up card” during the matchmaking process, they are not commonly used and are often forgotten.
Tantan
1. Matching Function Analysis
a) LBS + Image
This matching method is simple, direct and also more efficient. After users view actual photos of different profiles on the matchmaking page, they can either swipe right to like or swipe left to pass. If two users like each other, they will form a match and will be able to start a chat.
Before the matchmaking process, users can set filters in terms of location and age. After a successful match, users can enter the matched user’s profile page by tapping on their profile picture in the chat list. Photos form the main bulk of a user’s profile page. This is in line with Tantan’s appearance-focused positioning.
b) Quick Chat
Tantan’s swipe matchmaking method includes offline matchmaking. Therefore, there is no guarantee that the person being matched is currently online. Users might need to leave a message wait for the match to reply. The quick chat function avoids this waiting process. Each user gets 10 quick chat matches per day, while SVIP can get 90 matches per month.
2. Summary
Tantan’s LBS-based matchmaking mechanism allows users to match other users near them and develop a relationship in real life.
The direct and efficient swipe matching method has laid the foundation of Tantan’s business model. However, the appearance-focused way of interaction has led to an unbalanced gender ratio. In comparison to male users, female users generally receive an overwhelming number of matches.
Jimu
1. Matching Function Analysis
a) Swipe Match
Jimu is positioned as the go-to platform for young people to socialize. Its interface appears comparatively trendy. This platform has a similar matching method as Tantan, but users can better customize their matchmaking criteria. Jimu’s users can set matchmaking filters on location, age, gender, interested topics, number of pictures and profile tags. When there is a successful match, the system will notify the users and encourage both parties to interact. Users can see their matching degrees on the chat page. There are also guiding questions for users to start chatting.
b) Quick Match
The quick match function supports both vocal and text matchmaking. Users can choose and edit the topics they would like to talk about before the process.
2. Summary
Jimu’s unique product positioning has attracted many users from the younger generations. However, there is still much that can be done to improve the overall user experience.
While the Quick Match function is nothing special, users can customize topics before the matchmaking process. This helps users to start conversations quickly.
Future development of Soul’s matchmaking process
Soul relies on algorithms to match users based on personality and interests. As the paring criteria look at how similar two users are, this method is not perfect. Apart from the accuracy of the personality tests, other matchmaking considerations are also not taken into account.
To improve matchmaking accuracy, Soul needs to learn from Tantan — specifically, Tantan’s filter function to improve its matchmaking mechanism. This allows for more customizable criteria and choices for users to find the match they want. Here we have two potential solutions to strengthen Soul’s matchmaking process.
1. Improve the filtering mechanism to reduce unfit matches
First, Soul can provide more alternatives for users to customize their ideal match. By providing users with more filter options when using the love hotline, users will receive a more personalized and suitable partner.
Second, Soul should work to improve the accuracy of its matchmaking function. This can be done by optimizing the profile tag function. Currently, users can only tap on “change batch” to see more tags instead of viewing them all at once. Some profile tags are also not attractive to the users. By improving the profile tag function, users will be incentivized to tag themselves more accurately, improving the overall matchmaking results. Another way is to introduce an evaluation mechanism that enhances the user profile. After the first matchmaking process, users can rate each other anonymously and choose labels for each other. This data can serve as additional information to match users in future.
2. Increase the usage of add-on functional cards during matching
As mentioned previously, the monetization of the current matchmaking process is not optimized. Allowing add-on cards to pop up when the page is loading can increase users’ willingness to purchase.
