With many women in my life currently looking for a significant other, I realized how scary the experience of online dating world is.
Many existing dating apps leave many users feeling frustrated and burnt out from the repetition of meeting people only to find out they were not the right ones.
In this project I set up to solve and recreate a dating app that was different and focused on creating more meaningful matches.
Ellen is a human factors specialist working in the medical tech space. After trying out all the different apps that are out there, she's really frustrated and starts to lose hope of ever meeting someone right
I researched more about metrics from the existing dating app space to see if I can get any further insights.
of all online dating app users
are millennials
Increase in in-app audio and video calls since the start of the pandemic
of women ages 18-34 felt online dating was NOT a safe way to meet people
Existing apps to meet people often focus on just meeting people but don’t address the safety of participants.
That lead me to my first initial objective.
In order to tackle this problem I started out by writing assumptions and evaluating which were important to validate. The key assumptions were ranked based on risk to the project and ease of validation. I then made sure that the riskiest and hardest to validate assumptions were either verified or reduced through the interview.
I created my main hypothesis that I wanted to test from user interviews.
I believe that women are searching for a better way to meet people online.
I will know this is true when I see improved user feedback on their perceived safety and quality of standards.
I know this is wrong when users rather utilize other methods to meet people.
Because I wanted to understand people’s inherent attitudes towards online dating and get qualitative data, I selected to do over the phone interviews to understand what the current online dating experience was like.
Interviewees were selected based on the following criteria. To simplify the the process, I decided to focus on heterosexual relationships as the first iteration of the app. All the different relationship types can be added in future iterations.
Here are all the interviewees I was able to reach out to:
From my interviews, almost all interviewees expressed that they felt safe meeting people online. This was unexpected and different than my initial hypothesis.
This lead me to pivot the problem space to create better connections. My new hypothesis is that through creating better suited connections - Ellen, my persona who wants a long term relationship, can reduce a lot of the frustrations during her journey.
By grouping similar related topics from the interviews, I was able to arrive at the following key insights.
From here, I decided to focus on solving Ellen's key pain points. To get a better understanding, I mapped her journey from onboarding all the way to connecting with someone and having that connection end.
Improve and create a curated profile recommendation to reduce the burden of choice
Suggest and recommend topics to discuss when meeting someone new - to help the conversation be more organic.
Help nudge/remind participants in a chat to respond, and when conversation is exhausted, create new prompts to keep the conversation going.
I dug deeper into further additional statistics to refine my user group. In an online resource,
I was able to find the top 3 countries that was willing to pay for online dating services.
From this I decided to refine my user group even more.
This would allow the app to tap into the Chinese market eventually - greatly increasing business opportunities.
To generate ideas for features, I reviewed main competitors which were identified through the interviews. This showed a gap in the problem space for a brand new dating app that had employed a better algorithm, but also focused more on the personality of profiles. By finding people who are closer in traits and personalities, their relationship can also grow organically - something the interviewees iterated over and over again.
Another important distinction that will be key to make not call messages matches. By calling conversation matches, it's creating unnecessary burden and pressure on both parties.
Here are some of the initial wireframes I sketched from my inspiration. Testing out some of the different chat views and profiles
Overall, I conducted two rounds of user testing, with 5 different users in each round of testing for a total of 10 different users.
The following are the tasks that I ask the users to conduct during the test.
For user testing, I asked the user to go through the registration process as a brand new user to the app.
One of the main insights during user testing was the confusion from some of the modals. I improved the modals by keeping and incorporating more relevant information like below.
This gives the context and setting of why the modal appeared
Explain what the user needs to do to resolve the notification
Majority of the users ignored the notifications and when a modal was tried, it was too intrusive to the overall experience.
To mitigate it I used a built in banner to notify users so they can check it in their own time.
In order to confirm and evaluate if the changes in the process and variations had a positive impact on the user experience, a qualitative test was conducted through google forms.
From the post testing survey results conducted there were still some opportunities to clear up the sign up flow, especially regarding how to to set various profile filters, and how to discuss that with Eros. Through the changes, there was clear positive impacts in the interaction experiences with Eros. It would be important to conduct further testing in the clarity and usability of the sign up flow until it is completed and statistically significant.
Utilizing an AI chat bot to learn the preferences of the user in order to
Most dating website focus only on looks. By utilizing prompts similar to "Hinge's" design, people will be able to see these prompts and better gauge who the other person is before deciding to chat with them. This helps reduce the negative feelings caused by lower quality matches.
Having the ability to designate detailed, custom preferences instead of a standard pick list.
It was only after the second user test, I realized creating a post testing survey to get quantitate results would be much better than just comparing task success. After the initial edit, the task success rate significantly increased; however, there were still areas of improvements and insights that was much easier to gleam and evaluate once they were compared side to side.
In the effort to create a “different” dating app experience, I tried to create unique layouts. This resulted in confused users during testing. In the end, I ended up redesigning on top of industry best practices and I could have gotten additional insights out of the user testing.
The most challenging parts in the design was determining how to choose the most important features for the app. Determining how to show these features and ideas without digressing from the main task flow and leaving additional features for future iterations.
In the second user test, many were still confused about the flow being incomplete. The task flow and design was revised and so additional contextual testing will required to ensure those changes have fixed these issues.
Due to time constraints, only the primary and secondary user flow was completed. However, there were still many tasks and features that would greatly improve Chattr's experience.