MishaBook a demo

Aug 14, 2026

Creative Kill Criteria You Can Write Down

Creative kill criteria are predetermined performance thresholds - expressed as ROAS, CPA, impression share, or engagement rate minimums - that automatically trigger creative pause or elimination when breached. They replace subjective judgment with documented decision rules.

Why Kill Criteria Matter

Ad accounts accumulate creative debt. A creative that underperforms for 14 days still runs on day 21 because no one documented when to stop it. Budget that should flow to winners instead sustains mediocre performers.

Kill criteria create accountability. They force teams to define success before launch, not rationalize performance after. They also reduce decision fatigue - the rule fires, the creative pauses, no meeting required.

Without criteria, kill decisions happen late or never. A creative running at 1.8x ROAS gets paused when a new winner hits 3.2x. By then, the underperformer has burned $8k in budget that could have accelerated the winner.

Core Kill Criteria by Channel

Kill criteria vary by channel, product type, and account maturity. A new DTC brand testing creatives uses different thresholds than an established account optimizing for efficiency.

  • Facebook / Instagram: Pause if ROAS falls below 1.5x after 500 conversions or 14 days (whichever comes first). Eliminate if CTR drops below 0.8% after 10k impressions.
  • Google Search: Pause if CPA exceeds target CPA by 40% after 50 conversions. Eliminate if impression share falls below 15% for 7 consecutive days.
  • TikTok: Pause if ROAS < 1.2x after 300 conversions or 21 days. Eliminate if CPM exceeds $15 and CTR < 1.2% after 50k impressions.
  • Email: Pause if unsubscribe rate exceeds 0.5% or click rate falls below 1.5% (segment-dependent). Eliminate if open rate drops 20% below list average after 3 sends.
  • Google Shopping: Pause if ROAS < 2.0x after 100 conversions. Eliminate if impression share < 20% and CPC exceeds category median by 25%.

Structuring Kill Criteria: The Decision Tree

Kill criteria work best as a tiered system. Not all underperformance warrants elimination - some creatives need time or budget adjustment.

Tier 1 (Pause): Creative hits a soft threshold. It stops spending but remains in the account. Trigger: ROAS 20% below target, after minimum sample size.

Tier 2 (Eliminate): Creative hits a hard threshold or fails to recover after pause. Trigger: ROAS 40% below target, or no improvement within 7 days of pause.

Tier 3 (Archive): Creative is removed from active rotation but preserved for historical analysis. Trigger: Eliminated creatives after 30 days of zero spend.

Sample decision rule: Pause a Facebook creative if ROAS < 1.8x after 400 conversions. Eliminate if it remains paused for 7 days with no improvement. Archive after 30 days eliminated.

Minimum Sample Sizes Before Kill

Firing a creative too early wastes winners. Firing too late wastes budget on losers. Minimum sample sizes prevent both.

Sample size depends on baseline conversion rate and acceptable error margin. A 2% conversion rate needs ~500 conversions to reach statistical significance (95% confidence, 20% margin of error). A 0.5% rate needs ~2000.

  • High-conversion products (>3% CVR): 300 - 500 conversions minimum before kill decision.
  • Mid-conversion products (1-3% CVR): 500 - 1000 conversions minimum.
  • Low-conversion products (<1% CVR): 1000 - 2000 conversions minimum, or 21 days (whichever comes first).
  • If budget is constrained: Use 14 days as a hard stop regardless of conversions. Pause underperformers and reallocate budget to test new creatives faster.
  • If budget is abundant: Wait for statistical significance. A creative at 1.9x ROAS with 200 conversions is not yet a loser.

Documenting Kill Criteria in Your Account

Kill criteria must be written down and accessible. Spreadsheets, Notion docs, or ad platform rules all work. The format matters less than consistency.

Create a single source of truth: A spreadsheet with columns for channel, metric, threshold, sample size, and owner. Update it quarterly or when strategy shifts.

Example row: Facebook Ads | ROAS | 1.5x | 500 conversions | Marketing Manager | Updated Jan 2025.

Automate where possible. Most ad platforms support rules that pause creatives at threshold. Set up alerts so the team knows when a creative hits pause criteria without manual daily review.

Common Mistakes in Kill Criteria

Thresholds too tight: Killing creatives at 1.2x ROAS when target is 2.0x means most creatives die before they scale. Thresholds should be 20 - 40% below target, not 50%+.

No sample size floor: Pausing a creative after 50 conversions wastes winners. Always require minimum conversions or days before any kill decision.

Ignoring creative fatigue: A creative at 2.0x ROAS but declining CTR week-over-week may need pause for refresh, not elimination. Build fatigue signals into criteria.

One-size-fits-all rules: A $50k/day account and a $500/day account need different thresholds. Scale criteria to account size and budget velocity.

No recovery window: A creative paused at 1.5x ROAS should get 7 days to improve before elimination. Instant kills miss creatives that need budget reallocation to perform.

Seasonal and Cohort Adjustments

Kill criteria should shift with context. A creative that underperforms in January may be a winner in November. Audience cohorts also matter.

Adjust thresholds for: Seasonal demand (holiday vs. off-season), new vs. returning customer cohorts, geographic regions, and device type. A mobile-only creative needs different ROAS targets than desktop.

Document seasonal overrides: If Q4 thresholds differ from Q1, write it down. Example: Q4 Facebook ROAS threshold is 1.2x (vs. 1.5x baseline) due to higher volume and lower margins.

Review and reset criteria quarterly. A threshold that worked in 2024 may not fit 2025 if product margins, CAC, or audience composition shifted.

Questions

FAQ

Should kill criteria be the same for all creatives in a campaign?

No. Creatives in different audiences, placements, or customer cohorts should have different thresholds. A creative targeting existing customers can operate at lower ROAS than one targeting cold traffic. Document thresholds by segment, not by campaign.

What if a creative is declining but hasn't hit kill criteria yet?

Use a secondary metric: declining CTR or engagement rate week-over-week. If ROAS is stable but CTR drops 25% in 7 days, pause for creative fatigue refresh even if ROAS threshold hasn't breached. This prevents creatives from dying slowly.

How often should kill criteria be reviewed?

Quarterly at minimum. Review after major product changes, pricing shifts, or seasonal transitions. If account ROAS or CAC shifts by >15%, recalibrate thresholds. Document the date and reason for any change.

Can kill criteria be automated in ad platforms?

Yes. Facebook, Google, and TikTok all support rules that pause or disable creatives at performance thresholds. Set up rules for your primary metrics (ROAS, CPA, CTR) and pair them with alerts so the team reviews before the creative fully pauses.

Want this on your account?

Thirty minutes. Bring the number that keeps you up.

Book a demo

More from the blog