"Winning" Davos depends on what you measure. Volume (how much coverage) and sentiment (how positive) tell different stories.
Two Ways to Win
Volume measures attention—who dominated the conversation. Sentiment measures tone—whether that attention was friendly or hostile.
They don't always align. Trump had massive volume but poor sentiment. Prabowo had excellent sentiment but modest volume. The real winners had both.
The Volume Leaders
Who got the most coverage (by quote count):
- 1.Donald Trump — 118,477 quotes (37% of all coverage)
- 2.Volodymyr Zelenskyy — 18,129 quotes
- 3.Emmanuel Macron — 15,371 quotes
- 4.Mark Carney — 14,296 quotes
- 5.Ursula von der Leyen — 11,345 quotes
Trump's dominance was overwhelming—6.5x more coverage than second place.
The Sentiment Leaders
Who got the best tone (% positive, among figures with 1,000+ quotes):
- 1.Prabowo Subianto — 100% positive (1,787 quotes)
- 2.Jensen Huang — 96.7% positive (1,799 quotes)
- 3.Ursula von der Leyen — 95.5% positive (11,345 quotes)
- 4.Mette Frederiksen — 93.3% positive (3,248 quotes)
- 5.Mark Carney — 89.7% positive (14,296 quotes)
Note on methodology: Sentiment is classified by an LLM model. Scores near 100% or 0% should be interpreted with caution—smaller samples and classifier edge cases can skew results.
The Real Winner: Ursula von der Leyen
The European Commission President had the best combination: 11,345 quotes (5th in volume) with 95.5% positive sentiment. That's consistent positivity at scale.
Mark Carney also performed well: 14,296 quotes with 89.7% positive.
The Sentiment Losers
Who got the worst tone (% negative):
- 1.Donald Trump — 47.4% negative (with only 11.2% positive)
- 2.Scott Bessent — 34% negative (defending tariff policy)
- 3.Vladimir Putin — 13.1% negative (69% neutral—mostly just referenced in context)
Trump's coverage was uniquely negative. Nearly half of all mentions were classified as hostile.
What This Means
The leaders who combined both volume and sentiment—von der Leyen, Carney, Frederiksen—shaped the narrative most effectively.