Why Are Thresholds the Most Important Part of a Rule?
Automated rules are only as smart as the thresholds behind them. A threshold that is too strict will pause winners, while a threshold that is too loose allows bad ads to drain budget for days.
The goal is to create thresholds that reflect your true business targets and the reality of ad volatility.
- Thresholds should reflect actual profit targets, not vanity metrics
- Different funnel stages require different benchmarks
- Data volume and time window affect reliability
How Should You Set CPA Thresholds?
CPA thresholds should map directly to your allowable customer acquisition cost. Start with your blended target and adjust for campaign type.
- Prospecting: 10-30% above blended CPA target
- Retargeting: 10-20% below blended CPA target
- New tests: Use higher CPA ceilings until data stabilizes
- Minimum data: 10+ conversions and 3-7 day window
What ROAS Thresholds Are Realistic for Automation?
ROAS thresholds should reflect margin, refund rates, and true payback period. Do not use last-click ROAS without adjusting for attribution windows.
- Target ROAS: Use contribution margin, not just revenue
- Scale rule: Trigger above 1.3-1.6x target ROAS
- Reduce rule: Trigger below 0.7-0.9x target ROAS
- Pause rule: Trigger below 0.5x target ROAS with minimum spend
Where Do CTR and CPM Thresholds Fit In?
CTR and CPM are early indicators, not final decision metrics. Use them to flag creative fatigue or auction shifts, but avoid major actions without conversion data.
- CTR: Use relative drops, like 30% below account average
- CPM: Alert when CPM spikes 25-40% in a short window
- Action scope: Reduce budget or alert, not pause
- Data minimum: Require 2,000+ impressions
How Do You Combine Multiple Metrics Safely?
Multi-metric rules are safer because they reduce false positives. Combining spend, CPA/ROAS, and time window conditions creates more reliable triggers.
- Example: Pause if spend > $200 and ROAS < 0.6x over 7 days
- Guardrail: Require at least 10 conversions before action
- Confirm: Include a frequency or CTR drop for fatigue signals
- Staging: Alert first, then reduce, then pause
How Often Should You Recalibrate Thresholds?
Thresholds should evolve with seasonality, creative quality, and market competition. Set a review cadence that is frequent enough to keep rules relevant but not so frequent that you chase noise.
- Weekly spot checks for obvious outliers
- Monthly updates tied to campaign-level performance changes
- Seasonal adjustments ahead of major sales events
- Document changes so thresholds do not drift without context
Frequently Asked Questions About Automated Rule Thresholds
Start with a notification-only CPA or ROAS threshold using a 7-day window and minimum conversion requirements. This lets you see trigger frequency before automation.
Use fixed numbers when your targets are stable and percent-based thresholds when performance fluctuates. Many teams use a hybrid: a fixed CPA cap with a percentage buffer.
They are helpful as early warning signals for creative fatigue, but should not be the only reason to pause ads without conversion data.
A common baseline is $100-$200 in spend and 10+ conversions, but adjust based on your AOV and conversion rates.
Yes. Segment thresholds by funnel stage, audience temperature, and objective so the rule logic matches the expected performance profile.