Inside the Latin Music Ecosystems

Executive Summary

Latin social dance music genres operate under fundamentally different market structures. Using a July 2025 snapshot of Spotify’s full catalogue metadata, I analyze follower bases, streaming momentum, artist distributions, and playlist dynamics across Salsa, Bachata, Cumbia, Merengue, and related subgenres. Three structural patterns emerge, and most importantly, the degree of attention concentration varies dramatically across genres:

Ecosystem-driven markets (Cumbia, Salsa) show broad participation across emerging and established artists, with more distributed playlist exposure.
Hit-driven markets (Bachata, to a lesser extent Merengue) exhibit extreme concentration, where a small fraction of tracks and artists dominate visibility.
Seasonal and culturally anchored genres (Forró) demonstrate large follower bases with periodic search spikes that don’t always align with streaming momentum snapshots.

Across genres, structural differences imply distinct commercial risk profiles, from venture-style breakout markets to diversified ecosystem plays. They also imply distinct commercial strategies, from release timing and playlist pitching to touring and artist development. Rather than asking which genre is “bigger,” the more important question in my mind is: how does each genre’s ecosystem function?

Background

In late 2025, a snapshot of Spotify’s full catalogue metadata — including artists, tracks, genres, and playlist associations — was leaked to the public. While it does not include time-series streaming data, it offers a rare structural view into how genres are labeled, followed, and distributed across the platform.

In this analysis, I focus on Latin social dance music (Salsa, Bachata, Cumbia, Merengue, and related subgenres) to better understand how these culturally-driven ecosystems differ across followers, streaming momentum, and playlist exposure. Rather than ranking genres, my goal is to understand how their underlying market structures differ.

Caveats

The dataset is a snapshot as of July 2025. It does not include time-series stream data, so we cannot observe growth trajectories directly.
The dataset did not have total streams for each track. Instead, Spotify has a metric called popularity, whose explicit definition is not fully disclosed. Spotify states it reflects a combination of total streams and recent streaming activity. Throughout this article, I refer to it as a Hotness Index to emphasize momentum rather than lifetime popularity.
Genres in Spotify are assigned at the artist level, not the track level. To analyze track-level dynamics, I mapped tracks to their artists’ genre tags. Some artists carry multiple genre labels (e.g., salsa and cumbia), introducing potential overlap. I constructed an overlap matrix to measure this effect and found it to be limited, with expected overlaps such as Salsa–Salsa Romántica and Kizomba–Zouk.
Although Bad Bunny has released Salsa-influenced songs, he is not tagged as salsa in Spotify’s taxonomy. Including him would materially enhance Salsa’s metrics. To avoid cherry-picking, I excluded him from the Salsa cohort.
I will not provide links to the dataset here, but with enough digging, you can find it.

That all said, let's begin -- and 5, 6, 7...!

1. Genre Discovery

Spotify contains thousands of micro-genres, so I had to be systematic about isolating "Latin social dance" music genres. Here's what I did:

Excluded genres with fewer than 1,000 artists to remove fringe classifications.
Identified relevant genres using known names, alternate spellings, and partial keyword matches.

Because genre labels are assigned at the artist level and an artist can be labeled as multiple genres, I measured overlap between genres to assess potential double-counting. The normalized overlap heatmap below shows that cross-labeling is relatively limited, with exceptions in expected pairings such as:

Salsa + Salsa Romántica
Kizomba + Zouk
Son Cubano + Timba

While some overlap exists, it's not substantial enough to materially distort the broader structural patterns observed later. And besides, I adjust for these overlaps later when I aggregate the genres into families.

Note: I know genres like Flamenco aren't Latin dances, but I was curious to see its data. Also, I tried other approaches to discovering Latin social dance genres, like creating clusters of meso-genres based on frequency of genre co-occurrence (how often genres were mentioned with each other for a given artist) as well as clustering on semantic similarity, but I decided to keep the approach simple for this write-up -- although the results of the other approaches were very interesting... one meso-genre that formed was "Melodic/Symphonic/Speed Metal" and it has the most connections to other meso-genres. Who knew?

ANY-way, now that I have our genres, let's start analyzing them...

2. Genre Popularity: Followers and Hotness

Total follower counts reveal embedded fan bases across genres:

Forró and Cumbia have a huge amount of followers, suggesting large accumulated fan bases. Note that if I included Bad Bunny's 93 million followers into Salsa, then Salsa would overtake Cumbia but still be behind Forró. However, follower count alone does not imply current momentum — as seen when comparing these numbers to the Hotness Index (a couple charts down).

We can aggregate the chart by grouping each genre by their family (which I constructed). Here I avoid double-counting by adjusting for artists that were labeled as multiple genres in the same family.

Now let's look at the strength of each genre's recent streams. Taking into account all the artists in each genre, here are their respective mean and median Hotness Index score (how many streams they're getting lately):

Note that these aren't raw counts of streams, but rather an index that reflects the recent streams relative to other genres in Spotify's catalogue. Cumbia and Salsa romantica artists show higher mean and median scores relative to their peers. Interestingly, Forró doesn't have as many recent streams as I expected given their massive follower count.

Also note that this index is on a scale of 0-100 (100 being the highest), meaning that these genres are somewhat niche in the Spotify ecosystem (Bad Bunny, the most popular artist on Spotify, has a score of 100). A large gap between mean and median suggests that a small number of breakout artists disproportionately drive streaming momentum. Genres with tighter mean–median spreads exhibit more evenly distributed attention across their artist base.

I can do the same family aggregation here (again, avoiding double-counting):

3. Artist Density: Followers vs. Hotness

From here, I reduced the number of genres to 10 to simplify the analysis; specifically, I looked at Bachata, Salsa and Salsa-related, Cumbia and Cumbia-related, and Merengue. I plot each genre's artists as dots based on 1) the number of followers they have and 2) their Hotness Index score. These density plots reveal structural differences between genres. Number of followers is on the x-axis (log-scaled), and recent stream strength is on the y-axis.

And again, we can aggregate this by family:

Cumbia and Salsa display broad ecosystems: artists span from low-follower, low-momentum tiers to highly followed and trending acts. This implies that those markets are diversified with both grassroots and mainstream presence, with multiple pathways to visibility.

By contrast, Bachata and Merengue appear more bottom-heavy, with dense clustering in the low-follower, low-momentum quadrant. In conjunction with playlist concentration data (see below), this suggests the primary difference between these genres may not be total audience size, but how attention is distributed within each ecosystem. Genres like Cumbia and Salsa support a deep middle tier of artists, while Bachata and Merengue see a disproportionate attention to a few stars (and emerging artists might face steeper visibility barriers).

Further, note the vertical line of dots on the y-axis in the sub-plot for each genre; these represent the streams that certain artists are getting even though they have very few followers. Salsa shows meaningful streaming activity even among artists with minimal followers, suggesting that new or niche entrants may still achieve streaming exposure without a large pre-existing fan base. It could also suggest that Salsa listeners are more responsive to discovery tools (playlists, recommendations) that surface new or niche songs.

I tried adding contours to the scatterplots above to quantify and visualize the concentration of artists across the space, but they wouldn't render for some reason.

4. Playlist Placements By Genre

The share of tracks appearing in at least one playlist varies significantly across genres:

Salsa Romántica exhibits the highest share of its tracks placed into playlists, indicating strong playlist compatibility and curator adoption.

Concentration analysis reveals even sharper differences:

The concentration analysis is striking: in Bachata, the top 1% of tracks account for 68% of the entire genre's total playlist placements! This concentration gap may be the most economically meaningful difference across genres... in an ecosystem like Bachata, visibility behaves like venture capital, where a few breakout tracks capture outsized returns. In more distributed ecosystems like Salsa and Cumbia, exposure is more evenly spread, reducing reliance on singular hits.

Finally, we can break down the distribution of playlist placements for each genre. Playlist placement count is on the x-axis, and number of tracks is on the y-axis (log-scaled).

This data reiterates that Bachata exhibits a long tail dominated by a few widely circulated tracks, whereas other genres display comparatively more balanced exposure across their catalogues.

Oh, and quick note: Bad Bunny's Salsa hit "BAILE INoLVIDABLE" (which I didn't include into Salsa) has 65,751 playlist placements, higher than any of the tracks across all the genres plotted above.

5. Bonus: Google Trends Comparisons

Out of curiosity, I looked at worldwide Google search trends for music topics for each genre (as indicated by Google) for the past 5 years.

Above we see Salsa music (in orange) has almost always been the highest in search interest, and is even trending up over the past 5 years.

Forró is second and shows pronounced seasonal search spikes in June — consistent with Brazil’s Festas Juninas celebrations. However, recall that the Spotify dataset was taken at July 2025 (right after the Festas Juninas), and yet Forró's Hotness Index score wasn't very high. So Forró’s large Spotify follower base does not translate into consistently elevated recent streams or sustained search interest. Possible explanations include:

Followers reflect accumulated fan base rather than active listening velocity.
Search behavior differs by language and region.
Search spikes may not align precisely with streaming snapshot timing.

Anyway, next I looked at Google trends for dance topics (as labeled by Google) for Salsa and Bachata (and I included "Swing" out of curiosity). I didn't include Forró dancing or Kizomba dancing because they both registered at/near 0 throughout the past 5 years.

Salsa dance (in purple) consistently leads over the last 5 years and sees a huge spike in searches in February 2026, very likely due to Bad Bunny's Salsa-infused performance at the Superbowl that month. Also notable is Bachata dance's surge in early 2023, perhaps due to a confluence of the world re-opening after COVID and major Bachata congresses in Spain.

Takeaways and Commercial Implications

Strategy in Latin dance music is not simply a matter of followers, but how attention is structured. Highly concentrated genres imply returns are extreme but skewed, while more distributed genres see steadier returns. The structural differences between genres suggest different strategic playbooks:

Ecosystem-Driven Markets (e.g., Salsa, Cumbia): Multiple artists contribute meaningfully to the ecosystem. I'd think that these ecosystems provide more consistent revenue streams across multiple artists. Potential strategic takeaways:

Invest in long-term artist development.
Encourage consistent release schedules over singular breakout attempts.
Build community-driven touring circuits (festivals, socials).
Leverage diversified mid-tier talent.

Hit-Driven Markets (e.g., Bachata): This is a winner-takes-most market. High risk, high reward. Some strategy pointers:

Breakout probability is lower, so marketing intensity needs to be higher (thus, higher risk).
Allocate marketing capital only toward high-conviction singles.
Prioritize playlist pitching and viral amplification. Success might depend on high-leverage placements rather than on gradual catalogue growth.
Structure tours around established headliners, and/or pair emerging artist tours as opening acts for established headliners.

Seasonal Genres (e.g., Forró):

Time releases ahead of predictable cultural peaks.
Align touring schedules with seasonal demand and cultural calendars.

Salsa Dance Implications:

Finally, since interest in Salsa dance is surging, my advice to those in the Salsa dance community are:

Lean into the momentum. Interest in Salsa dance is at a rare, historic high amongst the general public.
Lower barriers to entry. New interest will likely come from a general audience that saw the Superbowl performance. So make your classes, events, and content accessible and friendly. Maybe even teach a simple sequence that was performed during the Superbowl.
Increase visibility now. Aggressively market your classes, organize and market more social events, engage in influencer partnerships.
Capture long-term retention. Figure out how to convert this influx of newcomers to repeat customers.

Extensions and Follow-on Topics

Some ideas for extending this analysis, some of which require different datasets:

Segment the above analysis by track release date.
Collaboration strategies across genres.
Live concert analysis: live events by genre; geographic distribution; festival clustering.
Return on investment/marketing spend demonstrated in streams, followers, concert sales.

Whew! This was my first deep-dive analysis into music. Reach out to me if you'd like to see this analysis for different genres, or have other ideas for analyzing the music space. It was fun -- thanks for reading!

-Daeil