Back to Projects
📊 Measurement & Attribution

Market Sentiment & Brand Perception

UCSD · Jan 2026

Built an NLP pipeline to analyze Twitter and digital news sentiment for multiple brands, discovering that trend-driven brands experienced faster and stronger short-term sentiment swings compared to established brands — a key insight for brand marketing teams managing reputation.

Key Insight

Trend-driven brands (hype-based, social-first) showed 2–3x faster sentiment velocity in response to events compared to established brands. This means brand teams for newer companies need faster response playbooks, while legacy brands have more buffer time but face stickier negative sentiment once it takes hold.

Tools & Technologies

PythonTwitter APINLTK/VADERPandasMatplotlib

Topics

NLPSentiment AnalysisBrand AnalyticsTwitter API
← All Projects