How AI Partners Transform Voice Search: Design Strategies for Emotional Interaction

Project Background

Voice‑search scenarios are evolving toward younger, more diversified experiences. An initial AI‑partner prototype was launched, but it failed to achieve emotionally interactive results, prompting a redesign focused on emotional connection and casual conversation.

Design Challenges & Goals

Key issues identified were:

After introducing the digital avatar, users clicked more and expressed chat desires, yet the avatar could not respond promptly, leaving emotional expressions unaddressed.

Users quickly left the voice‑search page without exploring additional features.

The redesign set three core objectives:

Build Partner Awareness: Strengthen recognition of the AI partner to stimulate voice‑search actions.

Satisfy Emotional Interaction: Enhance the emotional experience within the voice‑search context.

Enhance Immersive Experience: Create a searchable, interactive, controllable, and customizable immersive voice‑search environment.

Design Measures

1. Framework Upgrade – Building Recognition

The old panel mixed the avatar with functional buttons, causing confusion and limiting extensibility. The new layout separates the avatar from the panel, retains the microphone button for physical interaction, and uses the avatar as the central interactive element, improving emotional feedback while keeping the framework scalable.

The redesign also reorganizes functions by priority, ensuring clear information hierarchy, non‑overlapping hot zones, and unobtrusive feedback areas. Post‑launch metrics showed increased click‑through, higher panel flow, and reduced exit rates.

2. Emotional Layer – Full‑Scene Experience Upgrade

Beyond structural changes, each voice‑search state received emotional enhancements. Visual cues such as avatar gestures (waving, heart‑sign) replace static images, and youthful language narrows the gap between user and AI partner, increasing exposure and conversion.

Interactive bubble prompts simulate the avatar’s activities (listening to music, reading, gaming), encouraging clicks and conversation. Updated guidance rules adapt to different scenarios, linking bubble actions to corresponding “negative‑first‑floor” chat scenes without breaking emotional continuity.

3. Perception Upgrade – Immersive Visual Atmosphere

The redesign emphasizes a youthful, emotional style using large‑area frosted‑glass textures, vibrant colors, and rounded shapes to convey a light, immersive feel. Visual hierarchy focuses on three aspects:

Texture: Semi‑transparent matte finishes create depth and a premium look.

Shape: Predominantly curved UI elements (panels, buttons, icons) convey a friendly, youthful vibe.

Color: A carefully selected palette aligns with the AI partner’s attire, ensuring harmony and visual comfort while supporting future feature expansion.

Color rules include defining a primary hue range, extracting dominant colors from the avatar’s outfits, calibrating hue, saturation, and brightness to maintain consistency, and establishing complementary accent colors for varied scenarios.

Conclusion

By building partner awareness, delivering emotional interaction, and crafting an immersive visual environment, the AI partner’s exposure increased, user perception of intelligent search improved, and both search and interaction metrics rose, demonstrating the effectiveness of a holistic, emotion‑driven redesign.