Skip to content

ICFTBA Proceedings (2025) — Guerrilla marketing case study of Liquid Death

Original study
Academic
Publisher
Proceedings of ICFTBA 2025 Symposium
Evidence:
Observed behaviour
Vendor interest:
None
Sample:
Single-brand qualitative case study of Liquid Death; YouTube ad view and comment counts plus 50 sampled comments from each advertisement, compared with selected Coca-Cola ads.
Methodology
Disclosed
Published
2025

Summary:

This paper, written by student authors for the Proceedings of the ICFTBA 2025 Symposium, looks at how small brands can use guerrilla marketing to break into crowded markets. Its single case is Liquid Death, the canned water brand, and its tactics: viral, street, ambient, experiential and ambush marketing.

The method is a qualitative case study of one brand. The authors pulled view and comment counts from Liquid Death’s YouTube ads and compared comment-to-view ratios with ads picked from Coca-Cola. They also sorted 50 comments per ad into positive, neutral and negative. The rest of the paper describes campaigns without a defined dataset.

Key findings:
• Liquid Death’s YouTube ads drew far more comments per view than the Coke ads chosen for comparison: 0.013% vs 0.0006% in one pair and 0.072% vs 0.0004% in another.
• 78% of sampled comments on one Liquid Death ad were positive.
• The authors argue that street stunts, a coffin-shaped cooler, dark humour, a heavy metal look and festival ambush tactics turned plain water into a lifestyle brand.
• Their main claim is that bold humour, shock, niche targeting and social sharing can build attention and loyalty without a big-brand budget.

Margin notes:
This is a student conference paper built on one brand, so the results may not apply to other brands or markets. The paper says little about how ads were picked, how comments were coded or how the Coke ads were matched, 50 comments per ad is a small sample, and YouTube engagement can be skewed by algorithms, bots and paid promotion. Parts of the discussion are the authors’ interpretation rather than data.

Cards from this research

Coming soon

This feature isn’t available yet.