Pause waste, scale winners: a framework for daily campaign management
Most media buyers check campaigns once a day. The best ones have a system for knowing exactly which campaigns to touch and what to do to them. Here's that system.
Campaign management isn't hard. But it is disciplined. The difference between a mediocre media buyer and an excellent one isn't intelligence — it's having a repeatable system and running it consistently, even when it's boring.
The core logic
| Quadrant | Condition | Action |
|---|---|---|
| Scale Winner | ROAS > threshold AND spend < daily budget cap | Increase budget 15–20% |
| Leave Alone | ROAS near threshold AND volume is adequate | No touch. Let it run. |
| Watch | ROAS trending toward threshold from above | Monitor. Set alert. No action yet. |
| Pause Waste | ROAS below threshold for 3+ days | Pause. Document. Analyze why. |
The most common mistake is touching campaigns too often. Every budget change resets Meta's learning phase. Restraint is a skill.
Setting your ROAS threshold
- 1.Start with your gross margin percentage. If your product costs $30 to make and sells for $100, you're at 70% gross margin.
- 2.Decide your acceptable paid acquisition ratio. Most DTC brands target spending no more than 30–40% of revenue on paid ads.
- 3.ROAS threshold = 1 ÷ (target ad spend as % of revenue). At 33%, that's 3.0x. At 25%, it's 4.0x.
- 4.Adjust for CAC payback. If you're in a high-LTV business and willing to be patient, you can drop the threshold.
The daily check: 10 minutes or less
> THE MAVRICK BRIEF
Want this kind of thing in your inbox once a week?
// Written personally by Brian
The 3-day rule
Never pause a campaign after one bad day. Meta's delivery algorithm has natural variance — a bad Sunday doesn't mean a bad campaign. Use a 3-day rolling average for all pause decisions.
- →3-day ROAS below threshold: pause candidate
- →3-day ROAS above threshold: scale candidate
- →Single-day outlier (either direction): note it, don't act
- →Trend changing direction: investigate the creative or audience
Scaling mechanics: the 20% rule
When you scale a budget, scale by no more than 20% at a time. Going from $500/day to $1,000/day in one move can trigger a new learning phase and tank performance temporarily.
Weekly: the portfolio review
- →What's the blended ROAS across all campaigns?
- →Is total spend trending toward monthly budget?
- →Are there channels or placements systematically outperforming?
- →What percentage of spend is in learning phase campaigns (should be under 20%)?
Stop pulling data. Start commanding Mavrick.
10 free missions. Connects to your accounts in minutes.
> THE MAVRICK BRIEF
What operators are actually installing this week.
One short email a week. Tool stack changes. Workflows operators just installed. Patterns from inside Mavrick's customer base. No theory, no hype, nothing you have to “implement later.”
// Written personally by Brian · One click to unsubscribe
Brian MacDonald
Brian MacDonald is the founder of Mavrick, the AI coworker for marketing teams. Previously ran SetupClaw.tech, an AI deployment service for SMBs. Read more about Brian and the mission.