Derma Chatbot

A friendly chatbot that helps you build your skincare routine

My Role

Project Lead

  • Task prioritization and guidance for project success

  • Providing feedback, critique, and collaboration support to the team

  • Ensuring task completion through effective communication and task management

Design Researcher

  • Developing and refining testing guides

  • Conducting user interviews and usability tests

  • Synthesizing insights from target audience feedback

Visual Designer

  • Crafting final client presentations

  • Enhancing team deliverables with visual design

The Problem

The idea for a skincare assistant was inspired by my own confusion over where to begin in my own skincare journey. I pitched this idea to my class and it was chosen by peers as one of the ideas four our class to build.

I had many questions; “Which products work best for me?” , “How often do I need to use each product?” , and “What order do I use the products I have?,”

I worked in a team of four to design a chatbot that would assist target users with creating and maintaining a skincare routine.

User Research

PHASE ONE

We needed to discover who this product would best be suited for, the features it would need to be a successful product, and what value it would provide to users.

I interviewed 3 out of the 9 selected participants, aged 23-39, each of which had varying experience and concern with skincare. With the market researcher, I synthesized the data we gathered to draw insights into what features should be the focus of our chatbot.

Research Goals

Identify

The target audience, who will use this product.

Explore

How users engage with skincare products, the factors influencing their product choices, and the frequency of their skincare routines.

Evaluate

Accessibility requirements and preferences of the target audience to ensure that the platform caters to a diverse user base.

Affinity maps

Pulling quotes from the interviews, the User Researcher and I conducted an intense research synthesis session using the white boarding tool, FigJam. Together, we created groups of similar ideas together, and created insights that would inspire further development.

    1. People that have existing skin concerns are more interested in skincare products and routines.

    2. People trust the recommendations of those who have similar skin tones and skin types to them

    3. People will stay loyal to a brand if one product from that brand is successful for them

      TARGET AUDIENCE: People tackling an existing skincare situation

User artifacts

A user persona was created to assist in focusing our development on a specific user. Keeping this user in mind while developing the chatbot’s personality and features helped us prevent scope creep. The persona was further developed by crafting a user journey. The journey focused on how our persona, Solana, would accomplish a specific goal. From this we ideated how our virtual companion could simplify this process.

PHASE TWO

Creating Derma

With our target audience in mind, I now needed to hammer out who Derma is.

I collaborated with the Content Designer to define Derma’s voice, personality, and it’s overall character that would appeal to our target audience.

Derma journey map - Ten conversation flows were created to pin down how our bot would handle the situations we discovered in the creation of our journey map.

Mapping Derma’s personality and voice

Personality

Using Carol Pearson’s 12 brand archetypes, I defined Derma’s top 3 archetypes as:

The Caregiver - 70%

  • Driven by compassion and the desire to help others

  • Desires to care for, protect, and nurture others

  • Wants to do things for others

The Sage - 20%

  • An expert, scholar, advisor, researcher, thinker, mentor or teacher

  • Have a desire not only to understand the world but also to share that understanding with others

  • Help guide their audience to find the answers that they need

The Hero - 10%

  • Achieving mastery over a field is their goal

  • They want to motivate and encourage others

  • Live by the creed of honor and constant growth

Voice and Tone

To communicate to the team how I expected Derma to sound when talking with users, I crafted the two following phrases:

We love helping people find skincare routines that work for them

We hate the confusion and stress that starting a new skincare routine can bring

Derma would be friendly, but not disrespectful. She is a voice of authority, but is never condescending.

PHASE THREE

Usability Testing

With Derma’s character in place, and an idea of how our audience would use it, the wireframes created by our Technical Designer needed to be tested.

Testing approach

During a team meeting, I expressed concerns that using the prototype for our usability test would limit the variety of user responses.

I crafted a Wizard of Oz test, where users were given scenarios to complete while we provided responses in Derma's voice behind the scenes. This test, conducted with my team, yielded valuable insights to enhance Derma's user interactions.

Testing Goals

Observe

Observe how participants interact with prototype

Test

Test conversational dialogue (if it’s understood as intended

Map

Map target user mental model

Idenitify

Identify changes to be made

3 Testing Scenarios

One - Tailored Recommendation

Receive personalized recommendations for a winter skincare routine.

Two - Ingredient Understanding

Learn more about a specific ingredient that caused redness.

Three - Influencer Suggestions

Locate and gather reviews from this specific influencer to help inform your decision.

Main insights we gleaned

Give people choices

Readily available information created a streamlined chat: Add Visuals, pricing, ingredients, reviews to product recommendations.

People prefer visuals

When asking a user to describe themselves or a problem, providing visual references leads to more accurate results..

PHASE FOUR

Applying insghts

Working with the Technical and UI Designer, I assisted in applying the insights gained from our testing to refine and improve features for our final prototype.

Essential Product Information

Providing essential product information relevant to the conversation with the user by offering ratings, price, brand, and images for recommended products when needed.

“When Derma recommended products to me, I’d like to see the reviews and ratings then, so I don’t have to ask.”

Providing Visual Aids

Assisting users to make accurate choices with skin tone swatches to help users find their best represented skin tone

“I think for the skin tone options, maybe having colors next to them. So I could just choose one versus trying to pick a couple.”

Skincare Connection

Building trust by connecting users with skincare influencers and reviewers who share their specific needs o the user to establish a connection

“It's personal to me, because it's another brown woman that understands my hair type and understands affordability.”

CONCLUSION

Key takeaways and learnings

Working in a team allowed us to leverage our skill sets and generate a wide variety of creative ideas. At times, our conflicting perspectives provided a challenge, threatening the scope of our project. I learned to be a positive force for ideation, but to keep the focus realistic for the short turnaround of this project.

We all came with our preconcieved notions of how the chatbot should function. By prioritizing user feedback gained from interviews, testing, and research we were able to design Derma to be engaging and user-friendly.

Looking forward, improving Derma's language to be more natural and conversational would further enhance user interaction,. Additionally, collaborating with healthcare professionals could significantly bolster the chatbot's credibility. And of course, much more testing, iteration, and research would be crucial before full development

Next

CapCut