Budgets and when to scale
You don't need a big budget. Most artists run on $50 to $500 total. The real question isn't a dollar figure, it's whether you're giving Meta enough to learn.
Give Meta enough to learn
Meta's system has a learning phase: it needs roughly 50 results (link clicks, or conversions if you have a pixel) within 7 days before it settles on who responds and stops guessing. That's Meta's own rule. So size your budget around it: run a few days, see what one result costs you, and make sure your weekly budget can reach about 50 of them. If you spend so little that you only get a handful of results, Meta never optimizes and the money is mostly wasted.
Steady, and hands off while it learns
A steady daily budget beats dumping it all in one day. Don't edit the ad or audience mid-learning, a significant change restarts the 7-day clock.
How to tell if it's working
There's no universal "good" number to chase, and the averages you see online are some agency's sample, not a law. Judge it against yourself:
- Is your cost per result dropping over the week as Meta learns?
- Is your click-through climbing compared to your last campaign?
- Most important: is Pulse moving? Streams and monthly listeners ticking up is the whole point, more than clicks.
When to pause, scale, or leave it
- Pause an ad that's run past its learning week with a high cost per result and no Pulse movement. Try a new creative or a tighter audience.
- Scale a winner by raising the budget gradually. A big jump restarts learning, so nudge it up over a few days, don't double it overnight.
- Leave a working campaign alone. If it's delivering, let it run.
When you're ready to plan bigger or split a budget across campaigns, the dashboard and Roadie go deeper.