How to Actually Build a Trending Songs Compilation Without Losing Your Mind
I used to spend three days putting together a trending songs compilation for a client's playlist drop. Now it takes me about an hour and a half, give or take, if the API responses are cooperative. The difference wasn't some magic script. It was learning where the data actually lives and what the dashboards don't tell you. Start by pulling from the sources that matter. Spotify's Top 50 charts, Apple Music's daily top 100, and Billboard Hot 100 data are your foundation. Don't bother aggregating from TikTok trends unless your compilation is specifically for short-form video content. The tracking latency on TikTok is too unpredictable and the chart positions shift too fast for anything requiring even moderate accuracy. Here's what most people miss: the streaming platforms don't publish their trending data in real time. Spotify updates its charts once per day, usually around 6 AM ET. Apple Music does the same but at a different cutoff time. If you're pulling both sources simultaneously and expecting them to align, you're going to get duplicates and mismatches because Track A might be at position 23 on Spotify and position 41 on Apple Music at the same moment in time, and that gap matters if your client cares about consistency across regions.
The practical workflow I use now goes like this. I run a Python script that hits the Spotify Web API for the Top 50 of the target market, then the Apple Music API for their daily top 100. I normalize the track identifiers through a cross-reference lookup against the MusicBrainz database, which resolves the duplicate problem I used to fix by hand. That step alone dropped my average compilation time from two hours down to roughly forty minutes for a standard 30-track output. I need to flag a specific problem I hit last month that almost cost me a contract. I was compiling a trending songs compilation for a European market push, and the data looked fine until I cross-referenced the release dates. Several tracks listed as "trending" were actually re-entries — songs that had charted six months prior, got pushed out, and were cycling back in due to a viral moment. Spotify's algorithm treats these as separate trending events, but they inflate your compilation with stale momentum. The workaround was checking each track's first chart appearance date against the current trending window and dropping any re-entry older than sixty days. It took about three extra minutes per track and eliminated the problem entirely. Here's a counter-intuitive detail that trips people up: the "viral 50" charts on Spotify and Apple Music are not indicative of broad popularity. They're engagement velocity metrics, which means a track with a small but intensely active fanbase can rank higher than a track with widespread casual listening. If your compilation is meant to represent what's genuinely popular rather than what's generating the most engagement-per-listener, exclude the viral charts entirely. I learned this the hard way when a client complained that their compilation sounded like it was curated for a niche audience instead of the mainstream listeners they were targeting. We spent forty-five minutes manually swapping out viral-charted tracks for ones that were actually climbing the regular charts.
Another thing nobody mentions about regional variation: a trending song in Brazil is almost never trending in Germany, even within the same week. The cultural and platform differences are too significant. When building a global compilation, don't just average the charts together. Weight them by your actual audience distribution. If sixty percent of your listeners are in North America, the US and Canadian charts should dominate your selection. If you split everything evenly, your compilation will feel off to anyone in your primary markets because it's packed with tracks they've never heard of. For the actual compilation assembly, I use a simple sorting method: rank every track by its average chart position across all included sources, apply a minimum streaming threshold to filter out tracks that are climbing but not yet substantial, and cap the total at whatever your client needs. Thirty tracks is the standard. Forty-five works if you're covering multiple regions. Beyond that and you're just listing every song that's had a moment recently, which isn't a compilation anymore. It's a dump. The metadata handling is where things get tedious. Different platforms format artist names differently. "Doja Cat" on Spotify might show as "DojaCat" or include featured artists in unexpected ways. Run everything through a normalization pass before you export. I wrote a quick regex-based cleaner that standardizes artist name formatting and removes parenthetical feat. credits for the master list, then keeps the full metadata in a separate field. It added ten minutes to the process but saved me from having to manually check forty tracks for formatting consistency.
Get the Full Details
Export formats matter too. CSV is fine for internal use, but if your client wants to import the compilation into a DAW or a music management tool, they'll need either an M3U playlist file or a JSON object with full track URIs. Spotify's web API returns URIs that work across their ecosystem. Apple Music uses a different identifier system. You'll need to maintain a mapping table between the two. I keep a running spreadsheet for this and refresh it weekly because the mappings drift as new tracks get added to both catalogs. When I'm done, I verify the compilation by actually playing through it in order. Not the full thirty tracks every time, but the first five and the last five, plus a random sample from the middle. This catches awkward genre jumps, tempo mismatches, and tracks that are trending for the wrong reasons. It takes eight minutes and prevents about half the complaints I used to get after delivery. One final note on limitations: this method doesn't work well for niche genres. If you're building a trending compilation for ambient drone or hardcore punk, the mainstream chart APIs return almost nothing relevant and the genre-specific charts have different update schedules and fewer data points. In those cases, I pull from RateYourMusic charts and cross-reference with Bandcamp's daily trending section. It's slower, less automated, and the data quality is lower, but it's what you're stuck with when you're outside the pop mainstream.
