2026-08-09
What to Measure
One row per release, in one sheet, filled at two fixed moments. That is the whole system, and it outperforms every dashboard because it is the only place your numbers sit together. A related workplace perspective is covered in Monitask’s time tracking software.
The nine
Filled at release:
Date. Obvious and constantly omitted.
Hours spent, in the five categories. For additional current music-industry context, see Last.fm.
Money spent — recording, artwork, distribution, advertising, manufacture.
Filled at your fixed horizon — pick twelve or eighteen months and never vary it:
Reach, whatever the platforms report.
Saves or adds.
Repeat listeners.
Completion rate.
Follow-on actions — tickets, physical sales, merchandise, list signups.
And income, from the royalty statements rather than from an estimate.
What each answers
Hours and money against income gives whether it paid.
Saves per thousand reached gives whether the audience was the right one.
Completion gives whether the track holds, independent of any campaign.
Follow-on per thousand gives whether a route from listening to acting exists at all.
And the whole row against the previous rows gives the only comparison that controls for you being you.
Why two moments and not continuous
Because continuous measurement changes behaviour and produces anxiety without producing decisions.
A release does not have a meaningful weekly number. Checking daily teaches you the shape of a decay curve you cannot influence.
Two entries per release, one at launch and one at the horizon. Twenty minutes total.
Three releases
One row tells you nothing. There is no comparison.
Two rows tell you which was better and not why.
Three rows start showing pattern — whether the higher-spend release paid more, whether the one with fewer promotion hours did worse, whether completion tracks with anything.
Which means the sheet costs an hour and pays out in the second year, and that is the honest timeline. Anybody promising insight from the first release is selling something.
What to leave out
Follower counts. They move for reasons unconnected to a release.
Anything you cannot get consistently. A metric available for one release and not the next breaks the comparison, which is the sheet's only function.
And anything a platform reports that you do not understand. If you cannot say what a number means, recording it adds noise.
The short version
- Nine numbers, one row per release, filled at two moments: three at launch and six at a fixed horizon
- Pick twelve or eighteen months for the horizon and never vary it, because a number without a consistent window compares nothing
- Hours and money against income answers whether it paid; saves per thousand answers whether the audience was right
- Continuous measurement changes behaviour and produces anxiety without decisions — twenty minutes per release is enough
- One row tells you nothing, two tell you which was better, three start showing pattern
- Leave out follower counts, anything you cannot get consistently, and any number you cannot explain