Brkaway
Designing an AI reverse video search tool that reduces creator search time for brands by 40%.
Introduction
Brands spend hours combing through social media apps to find content creators that can produce content that’s aligned to their branding. This delays their social media posting schedule, leading to inconsistent posting, missed opportunities for timely content, reduced audience engagement, and can hinder their social media presence.
Outcome
I worked closely with the engineering team to understand tech requirements. I researched, ideated, iterated, created high fidelity mocks, prototyped, conducted usability tests, and delivered a comprehensive reverse video search tool that reduces a brand’s search time for relevant creators.
Role
Lead product designer
Timeline
4 months
Team
1 designer, 3 SWEs, 2 founders
Problem
Searching for content creators that can produce brand-aligned content is time consuming.
Brands spend hours searching on social media platforms like TikTok, Instagram, and YouTube to find creators who can produce content that is aligned with their branding.

HMW make it faster and easier for brands to find content creators that can produce content aligned with their branding?
Exploring & Ideating
Going broad, then narrow.
Starting with stickies, I led brainstorming session with the cross-functional team to ideate on different ways we could make the search process easier for brands.
Some standout ideas included:
Advanced search & filter
Allow brands to search and filter Brkaway’s catalog of creators. Filter by style, tone, industry, audience demographics, content types.
AI-powered matchmaking
Use machine learning to match brands with creators whose content style aligns with brand’s identity.
Dynamic gallery
Offer a dynamic gallery where brands can explore different content styles and save preferences or shortlist creators/content.
Consultation chatbot
AI-driven chat support that answers questions and can suggest creators in real time.
At the end of the brainstorming session, we unanimously decided to move forward with AI-powered matchmaking because it'd be an innovative solution that provides users with targeted and relevant search results.
Early Designs
Multiple rounds of low-fidelity designs, leading to the initial final designs.
Creating a series of low-fidelity designs allowed me to test out various layouts and ideas quickly.
Review & Refine
Many, many design crits.
I ran several design crits with leadership and cross-functional partners, which led to tons of iteration based on the great feedback received.
Upload page
To facilitate the reverse video search, users need a place where they can upload or link to the video they want to search.
Before
Relies on users uploading a video from desktop/mobile. Doesn’t support drag & drop or allow users to paste a video link.
After
Supports drag & drop, uploading directly from desktop/mobile, and allows users to paste a video link.
Loading states
A loading screen needs to be surfaced while the AI is working.
Before (V1)
Loading message moves too quickly after the analysis is done so users don’t see the rest of the messaging. Also not descriptive enough.
Before (V2)
Too much information presented and horizontal layout doesn’t work well in mobile.
After
I removed any unnecessary flavour text and aligned the steps vertically, making it easier to digest for users and mobile-friendly.
Showing matches
Users need to easily see the videos and creators that match their uploaded video.
Before
Seeing multiple creator videos like this wasn't ideal, and there wasn't enough information about each creator.
After
Creators have dedicated cards which display multiple matched videos, if any. More creator information is also provided.
Final Designs
Putting it all together.
The final result was an AI reverse video search tool that allows brands to upload content they like and find creators on our platform that can produce similar content.
© 2024 Amanda Sim. All Rights Reserved.
Made with 💖 and lots of brown sugar milk tea 🧋(30% sugar, less ice).
Last updated: Dec 2024