By

Karina Gandola

Published on December 22, 2025

For beauty marketers operating in the creator economy, influencer discovery has long been a time-intensive exercise shaped by instinct, manual filtering, and surface-level metrics. WHOTAG is betting that artificial intelligence, trained not just on engagement data, but on culture, aesthetics, and behavior, can change that equation.

At the center of that effort is Hannah Baek, Chief Strategy Officer at WHOTAG and VAIV, the South Korean-based AI and big data company behind the platform. Hannah describes her work simply: “I design technology that helps brands confidently choose influencers who truly make sense for them.”

WHOTAG officially launched globally in November 2025 after a beta phase across 109 countries, offering AI-powered influencer discovery to brands and agencies worldwide. The platform allows marketers to describe the type of influencer they want in any language and instantly receive recommendations based on aesthetic alignment, cultural context, content behavior, and audience signals.

Rather than filtering creators by follower counts or hashtags, WHOTAG applies what it calls GPT Profiling, a multimodal AI system trained specifically on beauty content and consumption patterns.

Hannah believes the timing is critical. “Influencer discovery in beauty was incredibly manual and inefficient,” she says. “We kept hearing the same complaint from marketers everywhere: ‘We spend so much time searching, and still end up with creators who don’t quite feel on-brand.’”

From Cultural Analytics to Creator Discovery

Hannah’s path to building WHOTAG did not begin in influencer marketing. She has spent nearly two decades working at the intersection of social big data, trend intelligence, and cultural analytics, analyzing how people express tastes, values, and identity online.

“For almost 20 years, I’ve analyzed how people express themselves online; their tastes, behaviors, aesthetics, values, and cultural signals,” Hannah explains. “That experience taught me one thing: numbers alone never tell the full story of influence. Context does.”

That insight became foundational to WHOTAG’s approach. “Traditional search tools rely heavily on follower count and engagement rate, which do not answer the core question: ‘Does this creator actually fit our brand’s tone and vibe?’” she says.

WHOTAG was designed to address this by teaching AI to interpret creators the way experienced marketers do: reading visual language, cultural nuance, and behavioral patterns embedded in content.

Image: WHOTAG Main Page

Why Beauty Became the Starting Point

WHOTAG launched with a deliberate focus on beauty, a decision Hannah says was driven by both data and industry demand.

“As we built our GPT Profiling technology, we found that more than 80% of our creator signals and content patterns came from beauty,” she says. “When we, as an official vendor of Meta, spoke with Meta, they confirmed the same insight – beauty is the largest and most dynamic category, and there isn’t even a close second.”

Beauty also presents a uniquely complex discovery challenge. According to Hannah, consumer expectations vary sharply across regions, and even shared terminology can mask cultural differences.

“Terms like ‘glow,’ ‘natural look,’ or ‘high coverage’ mean very different things across regions,” she explains. “That makes the limitations of existing tools especially painful for beauty marketers.”

By starting with what Hannah describes as the “hardest, most nuanced problem,” WHOTAG aims to build a foundation that can later expand into additional verticals. “Beauty is just the beginning of a much broader roadmap,” she points out.

How GPT Profiling Works in Practice

At the core of WHOTAG is its natural-language discovery interface. Instead of filtering through dashboards, users describe what they want in plain language (for example, “Hong Kong-based skincare micro-creators posting minimalist routines”) and receive results within seconds.

Image: WHOTAG Profiling Page

“WHOTAG is the fastest way for beauty marketers to find brand-fit influencers anywhere in the world,” Hannah says. “You describe exactly what you’re looking for, and WHOTAG instantly surfaces creators who match not just by hashtags, but by aesthetic, cultural vibe, and content behavior.”

Behind that is a multimodal AI system that analyzes both text and visuals. “WHOTAG leverages multimodal AI to process combined text and visual data,” Hannah explains. “Our GPT Profiling engine analyzes aesthetic signals and cultural context for precise brand matching.”

Rather than forcing marketers to adapt to rigid filters, Hannah says the platform adapts to the marketer. “You describe what you want in any language, and the model instantly interprets the cultural and visual context to return relevant creators.”

‘Brand Fit’ in Influencer Marketing

Hannah argues that signals are often invisible to conventional tools. “In beauty, true brand-fit comes from signals that go beyond basic metrics,” she says. “Content aesthetic, skin type or skin concerns, product affinity, audience reactions by region, cultural background. All of these influence whether a creator is the right match.”

WHOTAG’s system evaluates those signals directly from content, even when creators do not explicitly label them. “WHOTAG’s GPT Profiling interprets these signals directly from the creator’s content, even if they never explicitly describe their skin tone, concerns, or cultural background in text,” Hannah explains.

The platform also assesses risk and alignment factors that brands increasingly care about. “We evaluate deeper fit indicators: whether they’re active in social-impact communities, whether there are ethical risks, how many brands they’re collaborating with, and which brands they’ve worked with before,” Hannah says.

Image: WHOTAG Bucket

What Global Testing Revealed About Influence

During its beta phase, WHOTAG analyzed creator ecosystems across 109 countries. One of the clearest takeaways, Hannah notes, was how misleading the concept of “global” influence can be.

“‘Global’ beauty is far more ‘local’ than most people assume,” she says. “In many markets, local mid-tier creators outperform big-name stars with far larger followings.”

As Hannah reveals, brands testing WHOTAG frequently discovered creators in markets they had not previously prioritized. “Many brands told us they discovered creators with a true brand fit in countries they had never even considered,” she shares. “It was a clear reminder that real influence often comes from unexpected places.”

For Hannah, these findings reinforced the importance of cultural intelligence. “Content norms shift dramatically across regions, even within the same beauty category,” she notes. “Those differences reinforced why culturally intelligent models are essential for cross-border discovery.”

From Discovery to Insight and Prediction

Beyond matching creators, WHOTAG functions as an analytics platform that surfaces behavioral and consumption insights in real time.

“Marketers can surface insights in seconds that would normally take hours or even days to piece together,” Hannah says. Those insights include “early signals of trend formation across regions,” “how audiences react to different content styles,” and “overlap between a creator’s followers and competitors’ customers.”

Hannah sees this as a stepping stone toward a predictive strategy, adding that “AI will help marketers forecast which creators will drive specific outcomes, which aesthetic trends will rise next season, and which regions will respond to particular product benefits.”

She emphasizes that this extends beyond influencer selection. “Today, marketers can ask questions like, ‘How do Latina consumers in the U.S. describe skin-tint texture?’ or ‘How do Russian toner-pad users express their skin concerns?’ WHOTAG can answer these instantly.”

Separating AI Substance From AI Marketing

As “AI-powered” becomes a ubiquitous label in influencer tech, Hannah is candid about what she believes the industry gets wrong.

“Most services that claim to be ‘AI-powered’ are essentially doing automated filtering or tagging,” she says. “But real AI isn’t about processing more data. It’s about interpreting data with context.”

WHOTAG’s differentiation, she argues, comes from combining multimodal analysis with domain expertise. “We combine GPT’s reasoning with prompts shaped by 20 years of beauty research expertise,” Hannah explains. “That allows the model to understand creators not just as influencers, but as beauty consumers.”

The risk of relying on outdated tools, she warns, is strategic blind spots. “Marketers will miss cultural mismatch risks, emerging global micro-trends, and creators whose real influence doesn’t show up in traditional metrics.”

A Broader Beauty Intelligence Platform

As WHOTAG scales globally, Hannah hopes to shift how brands think about creator partnerships.

“We want to shift the industry conversation from ‘Who’s the most famous?’ to ‘Who can represent our brand most convincingly in this market?’” she says.

Her long-term vision extends beyond influencer discovery. “We’re working toward expanding WHOTAG into a broader Beauty Intelligence platform that not only finds creators, but helps brands plan, predict, and localize strategies with cultural precision.”

For Hannah, the goal is pragmatic rather than aspirational. “I want people working in beauty, especially in beauty marketing, to save time and drive better results by leveraging AI and data,” she says.

“Beauty influencer discovery is the starting point,” she concludes. “Not the finish line.”

Image source: WHOTAG