Publicis Media

Publicis Media is a global agency specializing in strategy, planning, buying, and data-driven solutions. This project explored how AI, crawling, and automated workflows could be used to identify early public attention signals and support more proactive Video & Audio media buying.

The media landscape within Video & Audio has become increasingly fragmented, with audience attention now spread across streaming platforms, social media, YouTube, podcasts, and online communities. Despite this shift, media planning and buying still rely heavily on historical data, benchmarks, and established planning models. The key insight behind the project was that cultural momentum and audience attention often appear in digital environments long before they become visible in traditional viewing or listening data. This created the opportunity to explore how AI, crawling, APIs, and automated workflows could be used to identify early public attention signals as a complementary decision-support layer for more proactive and relevant media buying.

The solution was built around an AI-driven workflow designed to identify early public attention signals connected to programs, sports events, podcasts, and media environments relevant for Video & Audio media buying. The workflow collects public data from platforms such as TikTok, YouTube, Reddit, and press pages using tools like Apify. AI is then used to analyze engagement, recurring themes, and signals of growing audience attention over time. For example, increasing discussion around a new TV series on TikTok before its release could indicate growing audience interest before it becomes visible in traditional viewing data. By identifying attention earlier, the solution could support more proactive and cost-efficient media buying, while increasing the potential to outperform existing benchmark levels before demand and competition fully increase.