Insight Mining

Insight mining is an AI feature that allows you to extract valuable insights from your selected research sources. You now have precise control over which sources are used and how insights are processed, with enhanced visibility into your data hub.

To Start

There are two ways to begin insight mining:

Option 1 - From an insight lane:

1. Create or select an insight lane in your journey

2. Click the AI icon within the insight lane

3. Select "Mine Insights" This pre-selects the lane as your destination for new insights

Option 2 - From journey level:

1. Click the Journey AI button in the top-right corner

2. Select "Mine Insights"This pre-selects all existing insight lanes as potential destinations

How It Works

The system processes your selected sources (up to 30 at once) from the embedded data hub to extract relevant quotes and create insights. The embedded data hub shows you:

• Source details: Title, type, quote count, and owner

• Last mined: When each source was last used in this specific journey

• Origin: Whether manually uploaded or from integrations (like Qualtrics → add link: https://www.theydo.com/help-center/qualtrics-workflows-for-ai-mining)

• Upload date: When the source was added to your workspace

Source Selection and Management

Find sources easily:

• Search by source name to locate specific files

• Filter by type (interview, survey, support log, feedback), owner, date range, or origin

• Configure the table view to show only the columns you need

Upload new sources: Click "Upload sources" within the mining flow to add new .txt or .csv files (up to 30 files, 50MB per selection) or paste text from other formats.

Configuration Options

Insight Types

Specify which types of insights to mine:

• Gains

• Pains

• Needs

• Observations

If no types are selected, the system will search for all insight types.

Destination Lanes

Choose which insight lanes should receive the new insights. You can select multiple lanes or change the pre-selected destination.

Mining Behavior

Control how AI processes your insights:

• Create only new insights - Generate fresh insights from your sources

• Only enrich existing insights - Add new quotes to current insights without creating new ones

• Do both - Enrich existing insights first, then create new ones (recommended)

How Insight Enrichment Works

The system intelligently manages your insights by:

• Prioritizing enrichment of existing insights with new supporting quotes before creating new insights

• Preventing duplicates by matching new content to existing insights first

• Maintaining consistency across your journey mapping

• Preserving context by linking quotes to appropriate insights

During Mining

• Destination lanes are locked during the mining process to prevent conflicts

• Other insight lanes remain accessible so you can continue working

• AI buttons are disabled in locked lanes until you accept or reject results

• Journey structure remains unchanged - AI only works within existing phases and steps

Best Practices

• Select relevant sources rather than mining everything - focus on quality over quantity

• Use specific insight types when you need targeted results

• Review the "last mined" column to avoid reprocessing recent sources unnecessarily

• Filter sources by date to focus on recent research or specific time periods

• Rate mining job quality after completion to help improve AI performance

• Review enriched insights before accepting to ensure accuracy and relevance