Overview

If a deal has a low win rate or isn’t delivering, the Maestro AI agent analyzes the possible causes. The agent examines the deal’s bid-funnel data, including the number of bids sent, accepted, and won. Based on this analysis, it identifies patterns that might indicate issues such as:

• Bids that are too low relative to the median CPM for the targeted inventory
• Limited available inventory
• Other configuration constraints
• Restrictive targeting

The agent structures each response into three parts:

• Findings: a summary of the relevant data.
• Causes: the most likely reasons for under-delivery or low win rates.
• Pricing misalignment: the agent can now compare your floor or bid price against the median CPM for the same inventory, giving you a concrete pricing reference rather than a general recommendation.
• Next steps: recommended actions in Maestro to improve performance, such as relaxing targeting, adjusting bids, or changing pacing.

This structure helps you review issues quickly, share insights, and determine which changes to apply.

Example prompts:

• Analyze why Deal-123 has a low win rate.
• Explain why meta-deal Meta-456 is not spending its daily budget.
• Why is Deal-123 not winning? Is my floor price competitive for this inventory?

Interact with Maestro AI agent

To troubleshoot delivery using bid performance analysis, follow these steps:

1. Select Ask AI in the top-right corner of the page and enter your prompt.
2. Provide any missing or unclear information if prompted. If the request is incomplete or ambiguous, the agent asks follow-up questions to gather the required details.