Working with Datasets in Scoop for Slack

Understanding Datasets

Datasets in Scoop for Slack come from three sources:

1. Organization Datasets - Shared company data
2. Personal Datasets - Your uploaded files
3. Channel Datasets - Data mapped to specific channels

![Screenshot: Dataset selector showing different dataset types]

Selecting a Dataset

Using the Dataset Command

In a channel:

@Scoop dataset

In a DM:

dataset

Alternative commands:

change dataset
switch dataset
list datasets

![Screenshot: Dataset selection dropdown interface]

Dataset Information

Each dataset shows:

• 📊 Name: Descriptive dataset title
• 📝 Description: What the data contains
• 🏢 Type: Organization, Personal, or Channel
• 📅 Last Updated: Data freshness

Organization Datasets

What Are They?

• Company-wide data sources
• Connected from your Scoop workspace
• Automatically refreshed
• Shared across team members

Common Examples:

• CRM data (Salesforce, HubSpot)
• Support tickets (Zendesk, Intercom)
• Marketing data (Google Analytics, Marketo)
• Financial data (QuickBooks, NetSuite)

Access Control

• Inherited from Scoop workspace permissions
• Administrators control availability
• Mapped to Slack workspace during setup

Personal Datasets

Creating Personal Datasets

1. Upload a CSV or Excel file
2. Scoop creates a personal dataset
3. Available only to you
4. Perfect for ad-hoc analysis

![Screenshot: Personal dataset created from uploaded file]

Managing Personal Datasets

• Automatically named after your file
• Stored in your personal workspace
• Can analyze immediately
• Switch between multiple uploads

Use Cases:

• Analyzing exported reports
• Combining external data
• Testing before sharing
• Personal productivity tracking

Channel-Specific Datasets

How They Work

• Administrators map datasets to channels
• Auto-selected when you join
• Ensures team alignment
• Reduces confusion

Benefits:

• Sales channel → Sales dataset
• Marketing channel → Marketing dataset
• No manual selection needed
• Team works with same data

Dataset Features

Automatic Intelligence

When you select a dataset, Scoop:

• Analyzes structure and content
• Generates semantic understanding
• Suggests relevant queries
• Identifies key metrics

![Screenshot: Dataset selected showing suggested queries]

Dataset Information Display

📊 Dataset: Customer Analytics
📋 Description: All customer data including transactions, support tickets, and engagement metrics
🔤 Columns: customer_id, name, revenue, tickets, last_activity, segment, ...

Smart Suggestions

Based on your dataset, Scoop suggests:

• Initial exploration queries
• Key metrics to track
• Potential insights to investigate
• Relevant visualizations

Switching Datasets

When to Switch

• Analyzing different business areas
• Comparing data sources
• Working on multiple projects
• Uploading new files

How Context Is Preserved

• Each dataset maintains separate context
• Chat history is dataset-specific
• Switch back to continue previous analysis
• No loss of work

![Screenshot: Switching between datasets maintaining context]

Best Practices

1. Name Datasets Clearly

For uploaded files, use descriptive names:

• ❌ data.csv
• ✅ Q4_2024_Sales_Report.csv

2. Verify Dataset Selection

Always check you're using the right dataset:

status

3. Understand Your Data

Ask Scoop about the dataset:

describe this dataset
what columns are available?
show me sample records

4. Use Appropriate Datasets

• Financial analysis → Financial dataset
• Customer insights → CRM dataset
• Campaign performance → Marketing dataset

Troubleshooting Datasets

Dataset Not Available?

• Check with your Scoop administrator
• Verify workspace mapping
• Ensure proper permissions

Can't See Expected Data?

• Confirm dataset selection
• Check data freshness
• Verify filters aren't applied

Need a New Dataset?

• Upload a file for immediate analysis
• Request addition from admin
• Check if already available in another form

Next: Uploading Files