Music Programmers Ponder the discrepancies between traditional “call-out” research and music streaming
Let’s delve into why these differences occur:
* Nature of Data Collection:
* Call-Out Research: In traditional call-out research, participants are asked to evaluate songs based on short clips or snippets. They provide feedback on whether they like or dislike the song, its catchiness, and other subjective aspects.
* Streaming Consumption Research: Streaming platforms collect data passively as users listen to full songs. This data includes play counts, skip rates, and user behavior over extended periods.
* Sampling Bias:
* Call-Out Research: Participants in call-out studies may not represent the entire listener population. Their preferences might not align with the broader audience.
* Streaming Consumption Research: Streaming data reflects actual behavior across diverse demographics, providing a more comprehensive view of what people actually listen to.
* Context and Intent:
* Call-Out Research: Participants evaluate songs in isolation, without considering real-world contexts (e.g., mood, activity, social setting).
* Streaming Consumption Research: Users stream music during various activities (commuting, working out, relaxing), which influences their choices. Streaming data captures this context.
* Sampling Duration:
* Call-Out Research: Participants hear short clips, often less than 30 seconds. Their preferences might change when listening to the full song.
* Streaming Consumption Research: Full-length song plays provide a more accurate representation of listener preferences.
* Recency Bias:
* Call-Out Research: Participants evaluate new songs, which may lead to a bias toward novelty.
* Streaming Consumption Research: Streaming data includes both new releases and older tracks, reflecting long-term popularity.
* Social Influence:
* Call-Out Research: Participants’ opinions may be influenced by perceived social norms or expectations.
* Streaming Consumption Research: Users choose songs independently, without external pressure.
* Algorithmic Recommendations:
* Streaming Consumption Research: Algorithms recommend songs based on user history, leading to personalized playlists. This influences consumption patterns.
* Call-Out Research: Doesn’t account for personalized recommendations.
* Radio Airplay and Repetition:
* Radio Stations: They often play songs repeatedly to maintain familiarity and audience retention. This practice can inflate airplay metrics.
* Streaming Services: User-driven choices determine play counts, reducing repetition bias.
* Monetary Incentives:
* Streaming Platforms: Artists earn royalties based on streams, incentivizing users to explore diverse content.
* Radio Stations: Airplay doesn’t directly impact artist earnings, so they may stick to safe, familiar songs.
* Industry Practices and Tradition:
* Radio: Legacy practices and programming traditions influence song selection.
* Streaming: Disrupts traditional models, allowing for greater diversity.
In summary, while call-out research provides valuable insights, streaming consumption data offers a more holistic view of music preferences. Radio stations, however, continue to balance tradition, audience expectations, and commercial interests when choosing songs for airplay12.