Reproduce “Agentic Personalisation of Cross-Channel Marketing Experiences” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Sami Abboud, Eleanor Hanna, Olivier Jeunen, Vineesha Raheja, Schaun Wheeler
Aampe developed an agentic framework for personalizing cross-channel marketing experiences, leveraging causal inference and contextual bandits with a modular action space to dynamically tailor messaging. Deployed to 150 million users, this framework achieved substantial increases in user intent, conversion rates, and double-digit relative Gross Merchandise Value in large-scale experiments.
Key insights
- A modular action space design effectively addresses the combinatorial explosion of possible message variations, enabling scalable personalization.
- Rigorous causal inference methods, like Difference-in-Differences, are crucial for accurately attributing engagement and conversion improvements to personalized interventions rather than mere correlations.
- Integrating human expertise through a 'human-in-the-loop' paradigm allows marketers to maintain brand consistency and quality control while benefiting from AI-driven optimization.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.