Meta Description

Evaluate and optimize meta-measures using three signals: coverage, content power, and audience match. Learn quality scoring rubrics, the 2×2 decision matrix, and when to archive, maintain, or expand ingredients for better targeting and conversion.

The Three-Signal Evaluation Framework

Remember: every ingredient is scored on three independent dimensions. To evaluate effectiveness, you look at all three together.

Signal 1: Coverage (How often does this ingredient appear?)

Coverage tells you what percentage of your contacts match this ingredient right now.

Coverage Ranges:

• Very Low Coverage (0–5%): Rare pattern. Examples: "CFO at cybersecurity companies" (only 3% of your contacts). Use when targeting a specific niche.
• Low Coverage (5–20%): Uncommon but present. Examples: "VP of Sales" (10% match rate). Good for selective outreach.
• Moderate Coverage (20–50%): Appears in about a quarter to half of your contacts. Examples: "Manager-level roles in SaaS" (35% match). Balanced for volume and specificity.
• High Coverage (50–75%): Matches most of your contacts. Examples: "Works at a company with 50+ employees" (65% match). Useful for broad categorization, less useful for targeting.
• Very High Coverage (75%+): Almost everyone matches. Example: "Interested in improving business efficiency" (88% match). Not useful for differentiation. Archive this ingredient.

The Coverage Check:

If an ingredient has very high coverage (75%+), it's not helping you separate good opportunities from bad ones. It's essentially noise. Archive it unless it pairs with other low-coverage ingredients to create a unique combination.

If an ingredient has very low coverage (0–5%), it might be too specific. But that's not always bad—it depends on Signal 2: Content Power.

Signal 2: Content Power (Does messaging about this ingredient resonate?)

Content power measures how much your messaging moves the needle for people matching this ingredient.

You measure content power by answering: When we send outreach mentioning benefits relevant to this ingredient, do people matching this ingredient respond better than people who don't?

High Content Power Example: