Optimize campaign performance with smart optimizations
Overview
This article explains how smart optimizations work, how recommendations are prioritized, and where to access them.
Understand smart optimizations
Smart optimizations is an intelligent operational assistant built directly into the Kamino platform. It scans your active line items every night for performance and delivery anomalies, evaluates pacing deltas, and delivers ready-to-apply recommendation cards across your dashboards.
Every smart optimization is delivered as a recommendation card structured around four core pillars:
- Diagnosis: a clear explanation of the detected anomaly, for example, under-pacing, restrictive targeting, or budget bottlenecks.
- Impact and metrics: contextual data showing the gap between expected and actual delivery, such as delivery lag or pacing delta.
- Corrective action: a concrete, pre-configured adjustment, for example, budget adjustments, pacing updates, or keyword or category expansion, that can be applied immediately.
- Feedback and status: real-time tracking of whether recommendations are pending, applied, refused, or expired.
RIC-based prioritization
Recommendations are ranked by their Reach, Impact, and Confidence (RIC) score to help you identify which issues to address first.
Each recommendation receives a criticality score from 0 to 100 and a corresponding urgency level:
| Level | Score range | Description |
|---|---|---|
| Critical | 66 or greater | High-urgency anomalies requiring immediate intervention (for example, severe under-delivery or imminent budget caps) |
| High | 44–65.9 | Significant issues that risk affecting campaign pacing or revenue |
| Medium | 16–43.9 | Moderate deviations or optimization opportunities |
| Low | Less than 16 | Minor findings or lower-priority adjustments |
The same score thresholds apply across campaigns, so recommendations with the same score receive the same urgency level.
RIC scoring formula
The criticality score is calculated as follows:
Criticality Score = Reach × Impact × Confidence × 100
The formula uses three parameters:
- Reach (0–1): the relative size of the finding compared with similar findings.
-
Impact (0.25–3.0): the estimated benefit of applying the recommendation, assigned according to the following scale:
- Minimal: 0.25
- Low: 0.5
- Medium: 1.0
- High: 2.0
- Massive: 3.0
- Confidence (0–1): the system’s certainty in the recommendation.
Access
You can access smart optimizations from the following locations:
- Optimizations: complete management view with full filtering across all statuses and historical action logs. To open it, select Optimizations from the left menu.
- Start your day: top critical recommendations prioritized to start your morning triage.
- Global dashboard: top 15 pending recommendation cards sorted by priority score for quick action.
- Dashboard for guaranteed campaigns: recommendations specific to the individual campaign and line items you are managing.
These views provide different levels of detail, from a prioritized daily overview to campaign-specific recommendations and recommendation history.