The four indicators
🟢 Opportunity
Performance is unusually strong. Worth investigating to understand what’s driving the improvement and whether it can be capitilised on.
⚠️ Caution
Performance is below typical levels but not critically so. Monitor closely over the next few days.
⛔️ Urgent
Performance is significantly below historical norms. Investigate immediately to identify and address the cause.
No indicator
Performance is within normal range. No action required. Note no visual indicator is actually displayed to limit cognitive load.
How thresholds are set
Different types of metrics use different sensitivity levels. This is because some metrics naturally fluctuate more than others, and we want to avoid both false alarms and missed issues. i.e. false positives & negatives.Volume metrics
Metrics like revenue, orders, and sessions are unbounded—they can range from zero to very large numbers and naturally fluctuate significantly. We use wider thresholds to avoid false alarms:| Indicator | Condition |
|---|---|
| ⛔️ Urgent | Below the 5th percentile |
| ⚠️ Caution | Between 5th and 15th percentile |
| 🟢 Opportunity | Above the 90th percentile |
| No indicator | Between 15th and 85th percentile |
Rate metrics
Metrics like conversion rate, CTR, and win rate are naturally bounded between 0% and 100%. Because they’re more stable, we use tighter thresholds:| Indicator | Condition |
|---|---|
| ⛔️ Urgent | Below the 10th percentile |
| ⚠️ Caution | Between 10th and 20th percentile |
| 🟢 Opportunity | Above the 85th percentile |
| No indicator | Between 20th and 80th percentile |
Efficiency metrics (lower is better)
For cost-based ratios like CAC, CPC, and CPM where lower values are better, the logic is inverted:| Indicator | Condition |
|---|---|
| ⛔️ Urgent | Above the 90th percentile (costs too high) |
| ⚠️ Caution | Between 80th and 90th percentile |
| 🟢 Opportunity | Below the 15th percentile (costs unusually low) |
| No indicator | Between 15th and 80th percentile |
Value metrics (higher is better)
For ratio metrics like AOV, LTV, and MER where higher is better:| Indicator | Condition |
|---|---|
| ⛔️ Urgent | Below the 5th percentile |
| ⚠️ Caution | Between 5th and 15th percentile |
| 🟢 Opportunity | Above the 90th percentile |
| No indicator | Between 15th and 85th percentile |
Responding to indicators
When you see ⛔️ Urgent
When you see ⛔️ Urgent
Act immediately. This metric is performing significantly worse than normal.
- Check if there’s an obvious cause (site issues, ad account problems, stock availability)
- Look at related metrics—if conversion rate dropped, did traffic source mix change?
- Compare against the same day last week to confirm the anomaly
- If you can’t identify a cause, escalate to your team
When you see ⚠️ Caution
When you see ⚠️ Caution
Monitor closely. Performance is below typical but not critical.
- Note the metric and check it again tomorrow
- If caution persists for 2-3 days, investigate as you would an urgent issue
- Consider whether external factors (seasonality, competitor activity) might explain it
When you see 🟢 Opportunity
When you see 🟢 Opportunity
Investigate and learn. Something is working unusually well.
- Identify what’s different—new creative, audience, or channel mix?
- Document the finding for your team
- Consider whether this can be replicated or scaled