Custom Reports in BrandWise — Builder, Presets, and Saved Views

Custom Reports Library

Custom reports let you create flexible data queries — from aggregated scenario tables to granular filters on individual responses. To open the library, navigate to Reports → Custom Reports in the project sidebar.

The library contains two report types:

Each report in the table shows: name, type (System / Custom), author, entity level, and last updated date. Custom reports can be copied, edited, duplicated, and deleted via the context menu.

8 System Presets

System presets are pre-built configurations covering typical analytics needs. They cannot be deleted, but you can duplicate them as a starting point for custom reports.

1. Scenario Health

• Level: scenario
• Shows: Overall Score, model count, evaluated responses for each scenario
• Use case: quickly compare scenario health

2. Model Effectiveness

• Level: model
• Shows: Overall Score + Mention Rate for each model
• Use case: understand which AI models best represent your brand

3. Competitor Mention Order

• Level: competitor_item
• Shows: competitor mention position and your brand's position in the same responses
• Use case: competitive analysis — who occupies your positions

4. Brand ToM Leaders

• Level: brand_tom
• Shows: mention count, Mention Rate, Top of Mind score, and delta relative to your brand
• Use case: understanding the top-of-mind landscape

5. Risky Responses

• Level: item (individual response)
• Shows: model, intent, query context, and Overall Score for responses where the brand is mentioned
• Filter: only responses with brand mention and low score
• Use case: finding responses where the brand appears in a negative or inaccurate context

6. Organic vs Prompted

• Level: item
• Shows: query context (Organic / Brand Prompted), Overall Score, Visibility, Relevance, Top of Mind
• Use case: comparing response quality when the brand is in the prompt vs. when the model recalls it independently

7. Sources

• Level: source_domain
• Shows: domain, model, citation count, average Overall Score, average Relevance and Usefulness
• Use case: understanding which web sources models cite when discussing your brand

8. Persona Performance

• Level: item
• Shows: persona, Overall Score, Visibility, Relevance, Top of Mind
• Use case: analyzing how different target audience profiles affect model responses

Report Builder

The builder lets you create a report from scratch or based on a system preset. The interface consists of a settings panel on the left and a live preview on the right.

Metadata

• Name — a descriptive report name for the library
• Description — optional explanation to help your team understand the report's purpose

Scope

Defines which data the report includes:

Dataset

The key builder section — defines the table structure:

Entity level — the data aggregation level:

Changing the entity level rebuilds the available columns.

Columns — select which data to include. Available columns depend on the entity level. Drag columns to reorder.

Sorting — choose a column and direction (ascending / descending).

Filters

A powerful filtering system lets you select exactly the data you need:

Filters are grouped: conditions within a group use AND logic, between groups — OR.

Results Preview

The right side of the builder shows a live table preview with current settings. The preview header displays the row count. This lets you verify the configuration before saving.

Keyboard Shortcuts

The builder supports undo/redo for quick rollback — use standard shortcuts (Ctrl+Z / Ctrl+Shift+Z).

Saved Views

Once saved, a report becomes available to the entire project team. Saved Views are named table configurations that you can:

• Create — any team member can save their own configuration
• Edit — update filters, columns, sorting
• Delete — remove when no longer needed
• Duplicate — create a copy for modification

Saved Views are especially useful for recurring analysis: configure a report once and return to it after each run without reconfiguring filters.

Creating Your First Custom Report

1. Navigate to Reports → Custom Reports
2. Click Create Report (or duplicate a system preset)
3. Enter name and description
4. Select period and scenarios in the Scope section
5. Choose entity level and desired columns
6. Add filters to narrow down data
7. Check the preview — make sure the table contains the right data
8. Click Save

The report will appear in the library and be accessible to your entire team.

Usage Examples

Find Weak Responses Mentioning the Brand

• Level: item
• Filter: "Mentioned" = Yes, Overall Score < 40
• Columns: intent, model, context, Overall Score, Visibility, Relevance
• Result: list of specific responses where the brand is poorly represented

Compare Models for a Specific Scenario

• Level: model
• Scope: one selected scenario
• Columns: model, Overall Score, Mention Rate, all 6 metrics
• Sort: by Overall Score, descending

Analyze Information Sources

• Level: source_domain
• Columns: domain, citation count, average score
• Sort: by citation count, descending
• Result: understanding which websites influence brand representation in AI

Next Steps

• Overview Dashboard — aggregated KPIs and trends
• Run Report — detailed analysis of a specific run
• BrandWise Metrics — formulas and components for all 6 metrics
• Create an account and get started