Blend Lab

Research Highlights

Nearly 40 Participants

Women aged 20–30 who regularly use social media for makeup inspiration and maintain established beauty routines.

Problem Statement

Beauty users are not lacking inspiration.

They lack understandable feedback that helps them recognize why a makeup style works—or does not work—for their own facial features.

Research Process

I conducted user interviews to learn more about how beauty users search for inspiration and evaluate makeup references.

Participants revealed that they often chose makeup looks from creators they were interested in and used those looks as references while applying their own makeup. However, during the actual process, they frequently found that the results on their own faces did not match their expectations.

你是否有稳定的化妆习惯？

每天0.0%

经常0.0%

偶尔0.0%

从不0.0%

你的妆容会参考流行/博主同款妆容吗？

经常0.0%

偶尔0.0%

从不0.0%

Key Insights

In the makeup selection process, users are not unwilling to experiment. What they lack is feedback that helps them understand why a look does or does not work for them.

Core Problem

Users' makeup experiences form a repeating loop:ExperimentLack of meaningful feedbackUnable to build judgmentContinue relying on external templates

Early low-fidelity sketches exploring makeup discovery, comparison, and personalization flows.

How Might We

How might we help beauty users understand why a makeup look works for their own facial features, so that each experiment builds reusable judgment rather than reinforcing dependence on external references?

Solution

User Flow

Core Features

Portrait used for mock AI facial feature analysis.

AI Face Analysis: Understand your facial features

The interactive analysis helps users understand their face shape, feature proportions, and beauty characteristics before exploring makeup adjustments. Instead of giving a fixed beauty answer, it turns AI feedback into readable cues for personal experimentation.

Blend Lab: Personalized makeup lab

Users can freely combine makeup elements and adjust styles, instead of having AI automatically match the “most suitable makeup,” to create their own aesthetic definition.

Community: A space for diverse aesthetic exchanges

Create a communication space in a non-traditional social media environment where users can see other users’ personalized aesthetic explorations, rather than the highly modified and optimized “standard beauty” on social media.

Design Value

VALUE 01

Active exploration over passive recommendation

Instead of continuously presenting users with predefined looks, Blend Lab gives them a lower-risk space to explore, compare, and combine makeup elements on their own.

VALUE 02

Reusable judgment over isolated results

Each experiment helps users understand how color, placement, proportion, and style relate to their own facial features, allowing experience to accumulate over time.

VALUE 03

Personal agency over external templates

The product supports users in forming a more confident and personal definition of beauty, rather than repeatedly returning to creators, trends, or algorithmic recommendations.

Interface Overview

Reflection

This project began as an exploration of AI-generated beauty experiences, but the research revealed a deeper problem: users did not simply need more inspiration. They needed a way to understand and learn from each attempt.

The most important design decision was to move away from automatically matching users with a supposedly “best” makeup look. Instead, I focused on creating an exploratory system that allows users to compare elements, adjust details, and gradually build judgment based on their own facial features.

If I continued developing Blend Lab, I would test how clearly users can connect specific makeup adjustments with visible outcomes. I would also explore how the product could provide lightweight feedback without replacing personal choice with another rigid recommendation system.

This project strengthened my understanding that AI-driven experiences should not only generate answers. They should help users understand the reasoning behind their choices and become more confident decision-makers over time.