Meta's official minimum is $1 per day, but that number is meaningless for music campaigns. At $1 to $5 per day, you will wait weeks to exit learning phase, and by the time you have data, your release window is gone.
The practical testing floor is $10 to $30 per day per creative. At that level, you can expect 1,000 to 3,000 impressions daily, enough ThruPlays to see whether your hook works, and a few intent signals (profile visits, link clicks, or saves) within the first 48 to 72 hours. Run at least 2 to 4 days before judging a creative.
Note Meta's learning phase historically targets about 50 optimization events in 7 days. If your budget only generates 3 to 5 events per day, the system stays stuck in learning and costs fluctuate unpredictably.
At $50 per day, you enter real testing territory: multiple creatives, faster iteration, and enough volume to compare hooks within a week. A common test-phase range is $15 to $60 per day per ad set, then scaling winners by 10% to 20% every few days once performance stabilizes.
The right question is not "what is the minimum?" but "how fast do I need to learn?" If you have 4 weeks before release, $15 per day works. If you have 7 days, you need $40 to $60 per day to compress the feedback loop.
Does $5 a Day Work for Music Ads?
At current CPMs (roughly $15 to $25 for music-focused Reels in competitive markets), $5 buys around 200 to 350 impressions daily -- enough to reach 1,000 to 2,000 people per week, but not enough for Meta to learn who actually engages with your music.
If your cost per save runs $2 to $4, $5 per day generates 1 to 2 saves. Meta needs roughly 50 optimization events in 7 days to exit learning phase. At that rate, you stay stuck in learning indefinitely: inconsistent delivery, no real optimization.
Tip $5/day can work as a maintenance budget to keep a proven winner running after you have already found what works at higher spend. It does not work as a testing budget.
When $5/day makes sense: You have already tested at $20 to $30 per day, found a winning creative, and want to keep it live at low cost while you focus budget elsewhere -- or you are running a long catalog play where you can wait months for signal to accumulate.
When it does not: You are testing new creatives, launching a release, or trying to learn what works. In those cases $15 to $30 per day is the practical floor, and even that is tight for campaigns optimizing toward saves or follows.