Skip to main content

Exporting Data from Scrunch

Breakdown of where to export data from Scrunch and what to expect in the results.

Scrunch includes several different export options across the platform, from a printable PDF snapshot of the Home page to flexible CSV/Excel exports from Explorer and metadata exports under Configure. Each export is designed for a specific use case, from quick sharing to deeper analysis in spreadsheets or BI tools. This article explains what each export does, what data it includes, and when to use it.

Looking for the exact Explorer name or API field for a metric? The Scrunch Metrics Reference lists every metric, breakdown and filter with the API field it maps to.


Home Page: Print to PDF

Where: Home page
Button: Use the page's print action (printer icon) to export a PDF snapshot

This export creates a PDF snapshot of exactly what you are viewing on the page.

What it includes:

  • All charts and tables currently visible

  • Any filters you have applied (date range, platform, persona, topic, etc.)

What it's best for:

  • Sharing a static report with stakeholders

  • Capturing a point-in-time view of performance

  • Presentations or internal updates

Important notes:

  • This is a visual export only

  • It is not intended for data analysis or manipulation


Explorer: Export Data (CSV or Excel)

Where: Explorer (under Observe) - build the metric, breakdown, and filter view you want on the canvas, then export it.
Button: The ⋮ menu on your Explorer chart or pivot, then Export data.

This export outputs whatever metric, breakdown, and filter combination you've built in Explorer. It is not a fixed schema. The columns you get reflect the metrics and breakdowns you selected, such as Mentions Rate, Position Average, Sentiment Average, or Citations Rate.

What it includes:

  • One row per combination of your chosen breakdowns (for example, prompt, platform, persona, geography, and time period)

  • Only the metrics currently added to your chart or pivot

  • Only the prompts and platforms currently visible based on your filters

How the export runs:

  • Exports run asynchronously. Scrunch compiles the underlying queries in the background and shows a download toast when the file is ready

  • Each export supports up to 8 queries

  • A pivot table with no Group by has no rows and cannot be exported. Add at least one breakdown before exporting

Sample columns (illustrative, based on what you've selected):

  • prompt

  • ai_platform

  • persona_name

  • country

  • date_week

  • Mentions Rate (API field: brand_presence_percentage)

  • Position Average (API field: brand_position_score)

  • Sentiment Average (API field: brand_sentiment_score)

  • Citations Rate (API field: brand_citation_rate)

Example row:

Are there any budget airlines that provide meals on long-haul flights? | Google AI Mode | Generic | US | 10/27/25 | 76.47% | 50 | 76 | 76.47%

What it's best for:

  • Weekly or monthly reporting

  • Trend analysis over time

  • Feeding dashboards or spreadsheets

  • Comparing performance by platform, persona, or geography


Explorer Citation Exports

Where: Explorer canvas, using a Citation Domain or Citation URL breakdown, then the ⋮ menu → Export data.

Citation exports come from Explorer builds, scoped to whatever metrics, breakdowns, and filters are in the chart or pivot you've created. Like any Explorer export, these run asynchronously (a download toast appears when the file is ready) and are capped at 8 queries per export.

Grok no longer contributes citations, so it will not appear in citation exports. Presence, ranking, and sentiment data from Grok are unaffected.

Citation Domain Breakdown (Summary)

Structure:

  • Build a Citation Domain breakdown in Explorer for one row per domain

  • Metrics aggregated across all prompts and time periods in scope

  • Export via the chart's ⋮ menu → Export data (CSV or Excel, delivered asynchronously via a download toast)

Best for:

  • Identifying which domains influence AI answers the most

  • Competitive benchmarking at the domain level

  • Executive or high-level reporting

Citation URL Breakdown (Detailed)

Structure:

  • Build a Citation URL breakdown in Explorer, grouped by date_week, for one row per URL and citing prompt, per week

  • Export via the chart's ⋮ menu → Export data

Best for:

  • Deep citation analysis

  • Understanding which prompts drive citations to specific URLs

  • Tracking how citation patterns change over time


Configure Exports: Prompts and Personas Exports

Where: Configure > Exports section
Downloads: Prompts (CSV) and Personas (CSV)

These exports are metadata exports, not performance exports. This section also lists Recent exports, which covers downloads requested from Explorer, dashboards, Citations, Prompts, and Site Maps.

Prompts Export

What it includes:

  • Prompt text

  • Stage

  • Persona name

  • Key topics

  • Tags

  • Branded status (Branded / Unbranded)

  • Platforms the prompt is monitoring

  • Prompt creation date

Go to Configure > Exports to download the Prompts export (CSV only).

Example:

Prompt | Stage | Persona Name | Key Topics | Tags | Branded | Platforms | Created Are there any budget airlines that provide meals on long-haul flights? | Evaluation | | Budget airlines, International Flights, Low-cost travel | | Unbranded | ChatGPT, Google AI Mode, Perplexity | 2024-03-15

Personas Export

What it includes:

  • Persona definitions

  • Associated metadata such as geolocation and descriptions

The Personas export is available as CSV only. To export, go to Configure > Exports and select Personas (CSV). All requested downloads also appear under Recent exports.

What these exports are best for:

  • Auditing prompt organization

  • Reviewing or updating metadata outside the UI

  • Joining with performance data for advanced analysis


Putting It All Together: Joining Metadata with Time-Series Data

Explorer now supports Tag and Topic as breakdowns directly alongside any metric, so you often don't need a separate metadata export just to slice performance by tag or topic. A join is only useful for metadata that isn't exposed as a breakdown, such as persona injection details.

If you need to combine that kind of metadata with performance data, you can still join two exports:

  1. Export Prompts metadata from Configure > Exports

  2. Export time-series performance data from Explorer (build the metric view, then use the ⋮ menu → Export data)

  3. Join the two datasets using prompt or prompt_id

This allows you to:

  • Analyze performance by metadata not available as an Explorer breakdown

  • Build custom reports not available directly in the UI

  • Create richer dashboards in Excel, Looker, or other BI tools


Quick Reference

Location

Export Type

Purpose

Home page

PDF

Visual snapshot with filters

Explorer canvas

CSV / Excel

Any metric, breakdown, and filter combination you've built (async, up to 8 queries)

Explorer (Citation Domain breakdown)

CSV / Excel

Domain-level citation impact

Explorer (Citation URL breakdown)

CSV / Excel

URL + prompt-level citation trends

Saved dashboard tile

CSV / Excel

That tile's underlying rows (pivot tiles export their leaderboard rollup)

Configure > Exports

Prompts (CSV)

Prompt metadata

Configure > Exports

Personas (CSV)

Persona metadata

CSV / Excel

Sitemap pages with audit scores and metrics

Did this answer your question?