← Back
EstablishedMajor 6 min read

The Catalog Audit

Finding the four songs in your back catalog actually doing the work — and the forty that aren't.

by · · 6 min read release-strategymarketing

A catalog audit is the once-a-year exercise of finding the four songs in your back catalog actually earning and the forty that aren't. At established tier most artists have rough intuition about which tracks "do well," but intuition is downstream of what the streaming app surfaces to them, which is downstream of what's already winning. The audit replaces intuition with a single number — save-rate per thousand streams — and re-ranks the fleet.

Score by save-rate, not by streams

Total streams measure exposure. Save-rate measures whether the listener wanted to come back. The two are uncorrelated more often than artists expect, and the gap is where the audit lives.

>2% Hidden gems Over-performs for its audience size. Push hard.
0.8-2% Workhorses The 4-8 songs doing 80% of the work. Feed annually.
<0.5% Graveyards Streams exist but intent is gone. Leave it.
Save-rate per 1k streams — the three buckets every catalog song falls into.

Total streams tell you what the algorithm decided to surface. Save-rate tells you what the audience decided to keep. Spend your budget on the second number.

The three buckets, in order

The catalog audit framework

  1. 01

    Pull the save-rate column

    Spotify for Artists exports per-song save-rate. Apple's equivalent is library-add rate. Run both, then average. The number to score against is 'saves per 1, 000 streams' — normalizes across catalog size.

  2. 02

    Sort into three buckets

    Hidden gems above 2%. Workhorses between 0.8% and 2%. Graveyards below 0.5%. The 0.5-0.8% band is a watch list — re-check in 90 days.

  3. 03

    Budget against the buckets

    Hidden gems get a release-style push (sync pitch, repackage, feature). Workhorses get a $500-$2, 000 anniversary refresh annually. Graveyards get nothing.

  4. 04

    Calendar the workhorses

    Assign each workhorse a refresh month so they're spaced across the year. The catalog now runs as a fleet, not a one-time decision.

  5. 05

    Re-run every January

    What was a graveyard at year three can be a hidden gem at year five. New listeners change the math. The audit is the re-rank.

Run this on every song over 12 months old. Score, sort, decide.

What the audit changes in the calendar

After the audit, the year has a structure. Two hidden gems get full release-style pushes (one in spring, one in fall). Six workhorses get refresh moments spaced across twelve months. The graveyards earn what they earn passively and stop pulling budget. The new singles you ship now have anchors in the catalog to pair with — the if you like this, hear this spine that the algorithm reads as cohesion.

The audit takes a Saturday. It shapes the budget for a year. Most established artists run it once and never look at their catalog the same way.

Frequently asked

What's a save-rate and why is it the right metric?
Save-rate is the number of unique saves divided by the number of streams. Total streams measure exposure; save-rate measures intent. A song with 2 million streams and a 0.4% save-rate is being algorithmically surfaced but not loved. A song with 200k streams and a 3% save-rate is loved by a real audience — that's the song to feed. Spotify for Artists exposes this number; Apple's equivalent is library-add rate.
How often should I run the audit?
Once a year, every January, on the calendar like taxes. Catalog behavior shifts as your audience grows — a song that was a graveyard in year three because nobody had found it can become a hidden gem in year five when 200k new listeners arrive. The audit is the moment you re-rank the fleet.
What does 'feeding a workhorse' actually mean?
Three moves — anniversary push (1, 5, 10-year), playlist refresh (re-pitch to mid-size editorial and curators), and a small visual refresh (lyric video, alt-cover, behind-the-scenes clip). Total budget per song: $500-$2,000. The workhorse already converts; you're giving the algorithm a reason to re-surface.
When is a hidden gem worth a real release-style push?
When save-rate per 1k streams sits above 2% and monthly stream velocity is flat or rising. That combination means the song over-performs and isn't dying — the algorithm has decided it's worth surfacing but the audience layer is too thin. A repackaging push (sync pitch, stripped version, feature add) can 5–10x its lifetime.
Should I delete the graveyards?
No. Deleting from streaming platforms breaks links and erases the small recurring revenue. Just stop spending on them. Don't pitch them, don't refresh them, don't include them in playlist plays. Let them earn what they earn passively. The active work goes to buckets one and two.