"Did the new single do better than the last one?" Most artists answer this by opening their stats and looking at total streams today. That compares eight months of an old song with three weeks of a new one, and the answer is almost always wrong. A fair comparison of two releases takes one spreadsheet, a few rules and about twenty minutes per release.
Count from release day, not from the calendar
The first rule is to compare equal windows measured from each song's own release date: day 1 to day 7, day 1 to day 28, day 1 to day 90. Never "September versus October", because one song might have been out for the whole month and the other for nine days.
Three windows are enough for most independent artists:
- The first 7 days show how the launch went: your existing audience, announcements, release day attention.
- The first 28 days show whether algorithmic features and playlists picked the song up.
- The first 90 days show whether it has a life beyond the launch.
Write the numbers down when each window closes. Some dashboards let you go back to any date range later, others do not, and figures sometimes get corrected afterwards. A note made on day 28 is a fixed point you can always return to.
Compare releases with the same source and metric
Compare like with like. Spotify streams for song A against Spotify streams for song B. Not Spotify streams for one against YouTube views for the other, and not "all platforms" for one when you only have Spotify data for the other.
The same goes for the type of number. Streams, listeners, saves and followers each measure something different. Pick a small set and use exactly the same set every time. If your daily statistics come from one place and your monthly earnings reports from another, do not mix them in one comparison: daily figures are early estimates, while reports arrive later and reflect what partners finally counted.
Keep a promotion log
Numbers do not explain themselves. Without notes, you will credit every difference to the song itself, when it might have come from a paid ad, a video, a collaboration or a holiday weekend.
For each release, write down briefly:
- what you posted and when, including short videos
- any ads you ran, with dates
- playlists that added the song, editorial or independent
- collaborations, features, interviews, live shows
- anything outside your control: a holiday, a big news event, a trend
Elvin Rza compared "Yay gecəsi" with his next single "Payız" and was disappointed: "Payız" had fewer streams in its first 28 days. His log reminded him that "Yay gecəsi" caught a short-video trend in week two, while "Payız" came out during exam season with almost no promotion. The comparison stopped being about which song was better and became a useful note about timing.
Look at ratios, not only totals
Totals reward reach. Ratios tell you how people responded once they heard the song. A few that are easy to calculate:
- Streams per listener. Higher means people came back to it.
- Saves per listener. Saving a song to a library is a strong sign of interest.
- Share from playlists versus your profile and libraries. Shows whether the song travelled beyond people who already know you.
- Followers gained during the window. Shows whether the song made people curious about you.
Ignore vanity metrics that look good and explain nothing: total impressions, likes on an announcement post, or a big stream count from one short placement that disappeared after three days.
One row per release
Make one row per release with the same columns:
- release date
- streams on day 7, day 28 and day 90
- listeners on day 28
- saves on day 28
- share of streams from playlists on day 28
- top three countries
- a short promotion summary from your log
After three or four releases the sheet starts to show patterns that no single screenshot can: which kind of song your audience saves, which promotion actually moved the numbers, which countries respond to which language.
Read the result honestly
With small numbers, small differences are noise. If one song had 900 streams in 28 days and the next one 1,050, that tells you almost nothing. Look for differences that are large and repeat across several releases.
Also remember that a new release often lifts older songs, because new listeners explore your profile. In the CULT dashboard, streams by platform and country appear for each reporting period, so you can check whether your whole catalogue grew around the new single, not just the single itself. Sometimes that is the real result of a release.