Project Overview:
Hooli users are not writing to each other anymore as they are relying on AI to write for them, our challenge was to have users truly communicate with each other and use our AI in a way we encourage them to express their thoughts and be human.
Problem Statement:
People are increasingly relying on technology to communicate, but this reliance can often lead to problems. People are not always saying what they really mean, and this can lead to misunderstandings. Additionally, people are often afraid to speak up and express their true feelings, preferring to rely on technology instead. This can be a problem because it means that people are not communicating effectively and may not be able to resolve conflicts. It is important for people to use technology responsibly, and to remember that it is only a tool for communication, not a replacement for human interaction.
Challenges:
- Have users truly communicate with each other and use our AI in a way we encourage them to express their thoughts and be human.
- Reduce The fear of not being able to express oneself authentically or of not being able to understand another person's point of view can be a barrier to communication.
- Using AI to help people overcome these fears and connect with each other on a deeper level.
Discover Phase:
Data overview
We looked at our data and find out the key terms that users were deleting from the final messages. From there we interview a set of 20 users to understand why.
Interview Insights
- Users are scared that the tone the person receives the message will be misunderstood rather than being afraid to say something.
- Users are lazy to defend their points through text and rather just drop it.
- Users forget to reply to specific messages because they are engaged in another activity when they receive the messages
Competitive Analysis
We knew that some competitors chat had better engagement so we took a quick dive to understand if we can find common patterns for their better engagmement. What we found is that most users rely on ‘brand awareness’ to consider an app more secure, free, or easier to use regardless of them actually being that.
Ideation Phase
For this experiment, as squad we decided to focus on Tone and Urgency. Below a list of the ideas we voted as majority to explore in our actual product.
Idea | Voting |
Provide user with “Feeling tone” UI to make sure individual get the right feeling. | ⭐⭐⭐⭐⭐⭐ |
Allow users to include a ‘timer for response’ mark so they get a response faster | ⭐⭐⭐ |
Use AI to refer to ethically reviewed sources when discussing topics | ⭐ |
Use AI to proof objectivity of thoughts and score | ⭐⭐ |
Enable AI from both sides to prompt a video or audio call when users seem to be spending to much time editing their messages | ⭐⭐⭐⭐⭐⭐ |
Create color codes for keywords that could be misinterpret by receiver based on their latest responses | ⭐⭐⭐⭐ |
Design Phase
Tone color
We explored bubble colors to enhance the tone the receiver meant and allow users to avoid a hostile conversation
Keyword coloring
Explore what keywords can look like when user is typing their message and provide with context as to why it may be misinterpreted
Prompt Video Feature
When users phone is in a public space, we will trigger the option for users to start a call or an automated message that say “let’s chat later”