agentleFS
Sign inStart free
Synter-Media-AISynter-Media-AI/free-skills/skills/kill-scale-rules/SKILL.md

kill-scale-rules

Skill4 starsChanged 6 months agoMITView on GitHub

Whether to kill or scale a campaign / ad set / creative. Covers kill thresholds with minimum-data guardrails and platform-specific scaling budget-step limits.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Scale in place: if the creative is still under its proven scaling ceiling and frequency is stable.

    line 46
  2. Distinctive:

    Duplicate to a new audience: if frequency is climbing or CPA rises after every budget step.

    line 47

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Kill Criteria & Scaling Rules

Most teams kill too early (acting on noise) or scale too aggressively (the algorithm can't keep up). These thresholds are calibrated so you act on signal, not noise.

Kill Criteria — minimum data required BEFORE killing

TimeframeKill ifMin data
24 hoursHook rate < 15%5K+ impressions
48 hoursCTR < 0.5%2K+ impressions
3 daysCPA > 2x target$50+ spend or 1K+ clicks
5 daysNo conversions3x target CPA in spend
7 daysCPA trending up 3 consecutive daysStatistically significant data
14 daysCPA 1.5x above target with no improvementFull test cycle complete

Never kill before minimum data. Bad decisions from small samples cost more than the extra test spend. If a creative has 800 impressions and 2 clicks, you don't have enough signal to conclude anything — wait.

Scaling Rules

ConditionActionFrequency
CPA < target for 48hIncrease budget 20%Every 2-3 days
CPA < 50% of target for 72hIncrease budget 30-50%Every 2 days
Winner holds after 3 increasesDuplicate to new audienceOnce per winner
Creative at 100K impressionsCommission 3 variationsImmediately
CPA rises after budget increaseRevert to previous budget, wait 48hAs needed

Platform-specific budget step limits

  • Meta: Max 30%/day budget increase. The algorithm needs recalibration time — bigger steps re-enter the learning phase and can wreck stable performance.
  • Google PMax: Tolerates up to 50%/day.
  • TikTok Smart+: Tolerates up to 50%/day.
  • LinkedIn: Conservative — 20%/day is the safe ceiling.

When to revert

If CPA rises >25% within 48h of a budget increase, revert to the previous budget and wait 48h before trying again. The algorithm may need to re-stabilize. Repeated failed scaling attempts on the same creative usually mean the audience is saturated — duplicate to a new audience instead of pushing harder.

When to duplicate vs. when to scale in place

  • Scale in place if the creative is still under its proven scaling ceiling and frequency is stable.
  • Duplicate to a new audience if frequency is climbing or CPA rises after every budget step.
  • Commission variations as soon as a creative crosses 100K impressions — fatigue is coming whether you can see it yet or not.

About this skill

Part of the Synter free skills collection — open-source agent skills for advertising, PPC, and marketing automation.

Want this skill (and 40+ more) running on autopilot against your live ad accounts? Try Synter — AI Agent Media Buyers that connect to Google, Meta, LinkedIn, TikTok, Reddit, Amazon, and 7+ more platforms.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free