In newsletter #3 we gave the honor to introduce another of our parners. This time the French podcast production company Louie Media:

Dear colleagues,

A quick word about us: Louie Media is a Paris-based podcast production company, launched by two journalists in March 2018, dedicated to creating narrative-driven audio content, with a strong editorial focus on societal issues, culture, and personal stories. 

Since the beginning, we have aimed to combine high production value with meaningful storytelling, building loyal audiences across multiple shows. Our most-listened podcasts are Émotions (26 175 882 listenings), Passages (19 046 593 listenings), Injustices. Our work sits at the intersection of journalism, creativity, and digital distribution, which naturally places data at the core of both our editorial and strategic decisions.

Our most successful data idea: measuring performance at D+30

One of the most impactful data approaches we’ve implemented is evaluating podcast episode performance at D+30 (30 days after release), rather than focusing primarily on immediate downloads.

In the podcast industry, there is often a strong emphasis on first-week or even first-day performance. However, we observed early on that our content – being largely narrative and evergreen – has a longer lifecycle. By shifting our primary KPI to listens at D+30, we gained a much more accurate view of an episode’s true reach and resonance.

This approach delivers value in several ways:

  • Editorial clarity: It allows us to assess which stories have lasting impact, rather than short-term spikes driven by promotion.
  • Better benchmarking: Comparing episodes at a consistent maturity point (30 days) removes noise and enables fairer comparisons across formats and shows.
  • Strategic alignment: It encourages teams to think beyond launch strategy and invest in long-tail discoverability (SEO, cross-promotion, catalog curation).
  • Advertiser relevance: It provides a more robust metric when discussing performance with partners, especially for campaigns that benefit from sustained listening.

Ultimately, D+30 has become a shared reference point across editorial, marketing, and business teams – helping us align on what “success” really means.

However, day-one performance also provides valuable insight: when an episode shows a high completion rate but relatively lower initial listens, it can indicate that the title may not be compelling enough. In that sense, D+1 metrics serve as a useful indicator of title effectiveness.

Completion rate is the most relevant metric to assess the editorial quality of a piece of content, as it allows us to identify precisely where listeners drop off and disengage.

A key challenge we still face

Despite these advances, one major challenge remains unresolved: understanding and attributing audience behavior across platforms and over time—while lacking precise data about who our listeners actually are.

Our podcasts are distributed across multiple listening platforms, each with its own analytics standards, delays, and limitations. While we can aggregate high-level metrics (such as total listens), we lack a unified, granular view of how listeners discover, engage with, and return to our content.

A key limitation is the lack of detailed, user-level data. Unlike other digital media, podcast analytics provide little insight into the people behind the listens. We often do not have access to precise demographic data, listening habits at the individual level, or reliable identifiers that would allow us to understand audience segments over time. In many ways, we know how much our content is consumed—but not by whom, nor in what context.

More specifically, we struggle to answer questions such as:

  • What truly drives long-term listening beyond the initial release window?
  • Who are our most engaged listeners, and how do their behaviors differ?
  • How do different distribution platforms contribute to audience growth versus retention?
  • To what extent do external factors (press coverage, social media, word-of-mouth) influence listening trajectories?
  • How can we measure the causal impact of paid marketing campaigns (e.g., Meta) on listening performance, given the limited access to platform APIs and the resulting gaps in attribution?

This fragmentation makes it difficult to build a coherent narrative around audience journeys. It also limits our ability to experiment rigorously, as we cannot fully trace cause-and-effect relationships between our actions and listener behavior.

Framed more broadly, the challenge is this: how can we move from aggregated performance metrics to a deeper, user-centric understanding of audio consumption—while working within a decentralized ecosystem that provides limited audience data?

We believe this is not just a technical issue, but a structural one that affects many publishers in the audio space. We hope that sharing this challenge can spark discussion, comparisons, and potentially collective solutions within this program.

What We Accomplished in Project CARL

  • Data-Focused E-Learning: Prensa Ibérica, Karjalainen and CVC have been developing an interactive e-learning course designed to bridge the gap between data analytics and everyday media work. It offers a practical guide to applying data insights in a newsroom context.
  • 22 New Case Studies: Under the coordination of Petit Press/sme.sk, the partnership has produced a comprehensive library of 22 case studies. These documents reflect the diverse needs of our partners and cover a broad range of strategic topics, including:
    • importance of video in podcasting;
    • comparing audience habits across different age segments;
    • implementing meta-tags for deeper content analysis;
    • balancing data metrics with editorial instinct;
    • content scoring;
    • and many more.

Note: All materials are available in English and all partner languages on this project website.

Next Steps in Project CARL

In the coming months, Louie Media will complete a comprehensive resource guide designed to help media professionals navigate the complex landscape of data-driven tools and methodologies.

To wrap up, all partners will begin organizing local dissemination conferences. These events will serve as a forum for sharing tips with peers and discussing the practical application of our findings with experts across the field.

We look forward to sharing more detailed schedules for these events shortly.