eSalon | Lead UX Designer
Hair Care Personalization
Defining the logic that maps customer inputs to personalized hair care recommendations.
The challenge
Personalization only works if people finish the quiz.
More questions can improve recommendations, but they also make people drop off. The issue was deciding which questions mattered, what the system could assume, and how to make each step feel like progress.
Halfway through the project, a first round of layoffs eliminated product management and UX research team members. I took over discovery, gathering input from executives in marketing, engineering, customer service, and R&D, turning it into the decisions that shaped a 0-to-1 experience.
My approach
Ask only the questions that matter.
I audited 19 personalization quizzes, including Function of Beauty and Prose. To learn which questions actually drove recommendations, I answered each quiz repeatedly with different answers and tracked how the recommended products changed. Most included questions that made no difference to the results. I cut similar questions from eSalon's quiz, including "How long is your hair?” and "After a wash, how long does it take for your hair to become oily?"
Question frequency by topic for each competitor analyzed.
One quiz, three kinds of customers.
New customers, returning customers who weren't logged in, and logged-in customers each arrived with different information available. Instead of sending everyone through the same steps, I designed the quiz to use existing profile data when possible. A logged-in customer with a completed hair color profile would start with 4 of the 8 steps already done, and no one has to log in or sign up just to begin.
Flow for handling system decisions and user types in the custom hair care quiz.
Build on what customers already trust.
Customers already trusted the hair color quiz, so I kept its familiar brand iconography and built the new quiz with eSalon's updated design system: tokens for color, typography, radius, and spacing; components like buttons and cards; and patterns like the multi-step percentage progress indicator and input fields.
Use prototypes to align stakeholders.
With no product manager or researcher left, I needed a way to get marketing, engineering, customer service, and R&D to agree on how recommendations should work. I built interactive Figma prototypes using variables and conditional logic that showed real branching and customized results, so stakeholders could use the quiz instead of reviewing static screens.
Outcome
The product team backed my research recommendations, along with the design and prototype, so I handed the design off to engineering. A second round of layoffs eliminated the remaining product and design team, including me, before the custom hair care line moved forward. With the team reduced by layoffs, engineering had limited resources and instead used content from R&D in the existing hair color quiz, so my design and recommendations haven’t shipped.
More of my work
-
![GoodRx app login flow for a recycled phone number, showing account matching via birthdate and SMS verification.]()
Recycled Mobile Number Verification
GoodRx | Senior Product Designer
-
![GoodRx Rewards experience showing how members can redeem points for prescription savings across desktop and mobile.]()
Point-of-Sale Rewards Redemption
GoodRx | Senior Product Designer
-
![GoodRx prescription savings card enrollment experience, helping members request a card and provide mailing information across desktop and mobile.]()
Prescription Savings Card
GoodRx | Senior Product Designer
-
![Colorsmith hair styling product catalog, helping customers browse and filter products across desktop and mobile.]()
Navigation & Product Discovery
eSalon (Colorsmith) | Lead UX Designer
