You posted three Reels, sent the song to a few blogs, played a small gig and boosted one post. A week later your streams are higher. Which of those things worked? If you did not write anything down, you will probably credit whatever felt most exciting, and next time you will spend your energy on a guess. Tracking which music promotion actually worked does not need special software. It needs a habit.
Keep a promotion log
The core tool is a log: one line for every action, written on the day you do it. A notes app works, a spreadsheet is better.
For each line, record:
- Date and time. Down to the hour for posts and ads, because timing matters when you read the numbers later.
- Channel. Instagram, TikTok, a Telegram channel, a radio interview, a blog, a gig.
- What exactly. Not just "posted a video" but "15-second clip of the chorus, filmed in the car, caption asking which line hits hardest".
- Who it reached. The country or city, and the kind of audience: your followers, a creator's followers, a radio region.
- Link used. Which link people were given, if any.
- Spend. Whether money was involved, so you can compare paid and free actions honestly.
Fill it in even when nothing happens. A post that changed nothing is useful information.
Give each channel its own link
When every channel points to the same link, you cannot tell which one brought people. The general fix is to give each channel its own way in.
- Separate links per channel. Many link services let you create several links to the same page and count clicks for each one. Use one for your bio, another for a blog, another for an ad.
- Tagged links. Some links can carry a small tag naming the source, so analytics tools can show where a visit came from.
- Different words for different places. On the radio you cannot give a link, but you can ask listeners to search for one specific phrase or song title, then watch whether searches rise.
- Codes for offline actions. For merch or tickets, a different code per channel shows where buyers came from.
Clicks are not streams, but they tell you which channel made people move.
Matching spikes with your actions
Now put the log next to your numbers. Look at daily streams where you can, and check which platform and which country moved.
A few rules make this more honest:
- Compare with a normal week. A rise only means something against your usual level.
- Expect a delay. A video can take days to reach people, and some platforms show data later than others.
- Look for location matches. If streams jumped in Türkiye, the Turkish blog is a better suspect than the gig in Baku.
- Check the platform's own sources. Artist tools on the big platforms often show whether listeners came from playlists, your profile, search or elsewhere. A jump from an algorithmic playlist has nothing to do with your post.
In the CULT dashboard, streams by platform and country appear each reporting period once partner reports are checked, which makes the country check easier over longer stretches.
Look beyond streams
Streams are only one signal. Also note:
- saves and adds to listeners' own playlists;
- new followers on the platform and on social media;
- comments, shares and direct messages;
- people who came to a gig and said where they heard you.
A paid ad that brings many plays but no saves or follows may have reached people who will never return. A small blog post that brings ten loyal listeners can be worth more over a year.
Do not fool yourself
The hardest part is admitting when two things happened at once. If you posted a video on the same day your track landed in a mix, you cannot know which one caused the spike.
Take Rauf Ay, a made-up rapper from Sumqayit. He sees a jump on a Wednesday and is sure his new TikTok did it. His log shows the TikTok went up on Tuesday evening, but also that a Turkish creator used his track the same night. Most new listeners are from Türkiye. On his next release he works with creators again on purpose, and the effect repeats. Now he has a pattern, not a hunch.
A monthly review
Once a month, read through the log and sort your actions into three groups: worth repeating, worth testing again and not worth the time. Drop one thing from the last group. Plan the next month around the first one. After a few releases you will be spending your limited time where it actually moves something.