From Pricing to Understanding

Panda Voiceover · Homepage Hero Redesign

The original Panda Voiceover homepage centered the experience around a word-count calculator. But client conversations revealed a different expectation: people wanted to describe the voice they imagined and let the platform help them find it.

That insight shaped a new direction for the homepage, combining natural-language search, structured talent data, and a simpler path from creative brief to relevant voice samples.

Product Research · UX/UI Design · NLP Requirements · Data Structuring

The original Panda Voiceover homepage led with a word-count calculator, allowing clients to estimate recording costs before exploring available voice talent.

While pricing was useful, the experience assumed clients already knew how to translate their creative needs into a service request.

The question became: What if we started with the voice they were imagining instead?

asked clients an open-ended question: If you could imagine the ideal voiceover platform, what would it look like?

Across conversations with ten clients, six expressed an expectation for a more AI-driven experience—one where they could describe their needs in natural language rather than navigate traditional search tools.

6 out of 10 clients expressed interest in prompt-based talent discovery.

The Design Opportunity

These conversations suggested that clients were beginning to expect more than filters and pricing calculators. They wanted a platform capable of understanding their creative briefs and helping them find relevant voice talent.

This became the foundation for exploring natural-language search, structured talent profiles, and a more intuitive discovery experience.

Instead of asking clients to navigate filters or calculate pricing first, the redesigned homepage introduces a natural-language search experience that lets them describe the voice they’re imagining—in their own words.

The idea goes beyond adding an AI-style search box. For the experience to work, the platform needs to translate what clients describe into information it can use to identify relevant voice talent.

We approached this through three connected components:

01 — Natural-Language Interpretation

Understanding requests such as “a warm, confident Saudi voice with a Hijazi accent” and identifying meaningful attributes like language, accent, tone, and intended use.

02 — Structured Talent Profiles

Organizing voice talent information into consistent, searchable attributes so the system can connect a client’s description with relevant profiles and voice samples.

03 — Matching & Filtering Logic

Defining how interpreted requests can be compared with talent attributes, narrowing the available options while keeping the discovery experience simple.

We also considered what happens when natural-language search doesn’t produce a suitable match. After three unsuccessful attempts, the proposed experience offers clients two alternatives: switch to traditional search filters or contact customer support for personalized assistance.

The goal is to make AI a helpful starting point, not a barrier to finding the right voice.