> ## Documentation Index
> Fetch the complete documentation index at: https://docs.wisepilot.app/llms.txt
> Use this file to discover all available pages before exploring further.

# Scoring

> Understand how WisePilot calculates optimization scores for your content assets.

Each content asset in WisePilot receives an optimization score — a weighted composite (0–100) that reflects how well it's performing against your defined objectives. Scores update automatically every night after the daily performance rollup completes — typically by early morning UTC. See [Automation & Data Freshness](/optimization/automation) for the full schedule.

## How Scores Are Calculated

The score for any asset is:

**Score = (Weighted Metric Sum) × Confidence Multiplier**

Where:

* **Weighted Metric Sum** — Each metric is normalized and multiplied by its configured weight
* **Confidence Multiplier** — A factor (0.0–1.0) based on data quality, applied to reduce scores when data is unreliable

A score of 75 on high confidence means "this asset is genuinely performing well." A score of 75 on low confidence means "early signals look good, but we need more data to be sure."

## Optimization Objectives

You can configure scoring rules for different optimization objectives. Each objective uses different metrics and weights:

| Objective           | Focus              | Key Metrics                                   |
| ------------------- | ------------------ | --------------------------------------------- |
| **Visibility**      | Being found        | Search impressions, rankings, organic traffic |
| **Engagement**      | Being consumed     | Pageviews, time on page, scroll depth         |
| **Offer Attention** | Being acted on     | CTA view rates, CTA click-through rates       |
| **Conversion**      | Generating results | Form submissions, lead attribution, revenue   |

### Example: Configuring a "Conversion" Objective

A conversion-focused scoring rule might use these weights:

| Metric                 | Weight   | Why                       |
| ---------------------- | -------- | ------------------------- |
| Form submissions       | 0.40     | Primary conversion signal |
| CTA click-through rate | 0.25     | Indicates offer relevance |
| CTA view rate          | 0.15     | Indicates CTA visibility  |
| Pageviews              | 0.10     | Traffic baseline          |
| Avg. time on page      | 0.10     | Engagement quality signal |
| **Total**              | **1.00** | Must sum to 1.0           |

This gives highest weight to actual conversions, moderate weight to CTA engagement, and low weight to traffic metrics.

### Example: Configuring a "Visibility" Objective

| Metric             | Weight   |
| ------------------ | -------- |
| Search impressions | 0.35     |
| Average position   | 0.30     |
| Organic clicks     | 0.20     |
| CTR                | 0.15     |
| **Total**          | **1.00** |

## Confidence Tiers

Not all scores are equally reliable. WisePilot assigns a confidence tier based on three data quality factors:

| Tier       | Badge  | Join Coverage | Freshness   | Sample Size   |
| ---------- | ------ | ------------- | ----------- | ------------- |
| **High**   | Green  | > 75%         | \< 24 hours | >= 100 events |
| **Medium** | Yellow | > 50%         | \< 48 hours | >= 50 events  |
| **Low**    | Red    | ≤ 50%         | > 48 hours  | \< 50 events  |

**What each factor means:**

* **Join coverage** — What % of events in the pipeline are successfully attributed? Low coverage means you're missing data. See [Data Quality](/optimization/data-quality).
* **Freshness** — How recently was data last collected? Stale data means the score may not reflect current reality.
* **Sample size** — How many events does this score draw from? Small samples are statistically unreliable.

<Note>Low-confidence scores are visually flagged in the UI with a red badge. Don't make optimization decisions based on low-confidence data — wait for more events to accumulate or fix the data quality issue first.</Note>

## Configuring Scoring Rules

1. Go to **Settings → Optimization → Scoring Rules**
2. Click **Create Rule** or edit an existing one
3. Select the objective (Visibility, Engagement, Offer Attention, or Conversion)
4. Set metric weights — use the sliders or enter values directly. They must sum to 1.0.
5. Set **priority thresholds**:
   * **High performer** — Score above this threshold is flagged green (e.g., > 70)
   * **Low performer** — Score below this threshold is flagged red (e.g., \< 30)
6. Save the rule

You can have multiple scoring rules active. Each asset shows its score for every active objective.

## Score Snapshots and Trends

Scores are saved as daily snapshots, enabling trend analysis:

* **Score trend chart** — See how an asset's score changes over time
* **Portfolio dashboard** — Aggregate score distribution across all assets
* **Delta column** — In the asset list, see the score change since last snapshot (↑ or ↓)

A steadily improving score indicates your content and optimization efforts are working. A declining score warrants investigation — check [Data Quality](/optimization/data-quality) first, then [Revision Impact](/reports/revision-impact).

<Tip>Scores reflect yesterday's data. If you published content today, expect meaningful scores to appear in 24–48 hours once enough events accumulate. See [Automation & Data Freshness](/optimization/automation) for details on data timing.</Tip>