Budget Decisions Made Simple: When to Scale, When to Kill

The hardest part of managing Meta Ads isn't creating campaigns—it's knowing what to do with them afterward. Should you increase budget? Decrease it? Wait? Give up?

Most advertisers make these decisions emotionally. A good day makes them optimistic. A bad day makes them panic. The result is constant second-guessing and whiplash changes that hurt performance.

Budget decisions should be mechanical. Set rules in advance, follow them consistently, and remove emotion from the equation. Here's how.

The Fundamental Principle

Every budget decision comes down to one question: is this campaign making money or losing it?

That seems obvious, but most advertisers struggle to answer it clearly. ROAS looks good one day, bad the next. Spend is up but so is CPA. The data is noisy and confusing.

The solution is to establish clear thresholds and time requirements before making any changes. Here's the framework.

When to Kill

Kill means pause the campaign completely. You're done spending money on it.

The key insight: don't kill too fast, but don't wait forever either.

If you kill after $20 in spend, you're not giving the algorithm time to optimize. You're making decisions on noise, not signal. But if you let a loser run for weeks hoping it'll turn around, you're burning money on wishful thinking.

The 2-3x CPA threshold gives you enough data to make a real judgment. If your target CPA is $30 and you've spent $75 with zero conversions, that's probably a bad campaign—not just bad luck.

When to Watch

Watch means maintain current budget but monitor closely. The campaign isn't bad enough to kill but isn't good enough to scale.

Most campaigns live in Watch for a while. That's fine. Not everything is clearly a winner or loser immediately.

The danger is leaving things in Watch forever. Set a time limit: if a campaign has been in Watch for 2-3 weeks without moving to Scale, either optimize it significantly or kill it. Eternal Watch is just slow losing.

When to Scale

Scale means increase budget. You want to capture more of this profitable traffic.

Consistency is the key word. One good day doesn't mean you should scale. Five good days in a row probably does.

The frequency check is important. If your frequency is already high, scaling budget will show ads to the same people even more often. That typically tanks performance. Scale when there's still room to reach new people.

How Much to Scale

When you decide to scale, don't double your budget overnight. The algorithm optimizes based on patterns, and sudden budget changes disrupt those patterns.

The 20-30% rule: Increase budget by 20-30% at a time, wait 3-5 days, then evaluate whether to increase again.

This feels slow. You found a winner and you want to pour money into it. But patience here protects you from the scaling crash—where a campaign that was doing great suddenly tanks when you increase spend too aggressively.

The algorithm needs time to find new pockets of your audience. Give it that time.

The Numbers You Need

To apply this framework, you need to know a few things about your business:

Your breakeven ROAS

This is the ROAS at which you make zero profit. Below this, you're losing money. Above it, you're making money.

Calculate: AOV / Contribution Margin = Breakeven ROAS

Your target ROAS

This is the ROAS you're trying to achieve. It should be meaningfully above breakeven—high enough that you're comfortable scaling aggressively.

Typical: 1.5-2x your breakeven ROAS

Your target CPA

The maximum cost per acquisition you're willing to pay. This is your profit margin divided by how much profit you need per sale.

Your minimum decision spend

How much a campaign needs to spend before you make judgments. Usually 2-3x your target CPA.

A Decision Tree

Here's the entire process as a simple flowchart:

1. Has the campaign spent your minimum threshold?
   • No → Wait. Don't make any decisions yet.
   • Yes → Continue to step 2
2. Is ROAS above your target?
   • Yes → Check if it's been consistent for 5+ days. If yes, Scale. If no, Watch.
   • No → Continue to step 3
3. Is ROAS above breakeven?
   • Yes → Watch. It's making money but not crushing it.
   • No → Continue to step 4
4. Has it been losing money for 5+ days?
   • Yes → Kill it.
   • No → Watch for now. Give it a few more days.

That's it. Four questions. Clear answers. No endless deliberation.

Common Mistakes

Scaling too fast

You double the budget after two good days and performance tanks. Slow down. The algorithm needs stability.

Killing too slow

You let a campaign bleed money for weeks because "maybe it'll turn around." It won't. Cut your losses.

Not having thresholds

You make decisions based on vibes instead of numbers. That's how you end up second-guessing everything.

Ignoring frequency

You scale a campaign with 3.5 frequency and wonder why it stops working. You exhausted the audience. Check this before scaling.

Looking at wrong timeframes

You panic over yesterday's results when you should be looking at 7-day trends. Daily fluctuations are normal. Trends matter.

Automating the Process

Once you have clear rules, you can automate much of this process. Meta offers automated rules that can pause campaigns or adjust budgets based on conditions.

But Meta's rules are limited and the interface is clunky. Better tools give you:

• Visual verdicts for every campaign (Scale/Watch/Kill)
• One-click actions to implement changes
• Alerts when campaigns cross thresholds
• Historical tracking of decisions made

The goal is to spend less time analyzing and more time acting. Clear verdicts based on your custom thresholds let you make decisions in seconds instead of hours.

The Bottom Line

Budget decisions aren't complicated. They just require discipline.

Set your thresholds before you need them. Follow them consistently when you do. Remove emotion from the equation by having mechanical rules for every situation.

Scale winners slowly. Kill losers quickly. Watch everything else until it becomes one or the other. That's the entire strategy.