Smart Columns

Smart Columns are AI-powered custom fields that extract specific information from LinkedIn profiles. Define what you want to know, and our AI will analyze profiles and return the data you need.

What Are Smart Columns?

Smart Columns use advanced AI to read and understand LinkedIn profiles, extracting exactly the information you specify. Unlike traditional enrichment that returns fixed fields (email, phone), Smart Columns let you ask for any custom data point.

How It Works

1. Create a Smart Column

Navigate to any lead list and click "Add Smart Column":

Name Your Column: Give it a descriptive name (e.g., "Years of Experience", "Tech Stack", "Budget Authority")
Define the Query: Describe what information you want extracted
Select Leads: Choose which leads to process (all leads, first 10, first 20, or custom count)
Start Processing: Click to begin AI analysis

2. AI Processing

Our AI system:

Fetches the LinkedIn profile data
Analyzes the profile content using advanced language models
Extracts the requested information
Validates and formats the response
Returns the processed value

3. Real-Time Updates

Progress updates via WebSocket connections
See results appear as they're processed
Status indicators show: Queued, Processing, Completed, Failed
Results are saved automatically

Use Cases

Sales Intelligence

Budget Authority

Query: "Does this person have budget authority? Return Yes/No"
Use Case: Qualify leads based on decision-making power

Company Size

Query: "What is the current employee count at their company?"
Use Case: Segment by company size for targeted outreach

Tech Stack

Query: "What technologies does this person mention using?"
Use Case: Identify prospects using specific tools

Recruiting

Years of Experience

Query: "How many years of total work experience does this person have?"
Use Case: Filter candidates by experience level

Skills Assessment

Query: "Does this person have Python and AWS experience? Return Yes/No"
Use Case: Technical skill matching

Management Experience

Query: "Has this person managed a team? If yes, what size?"
Use Case: Leadership role qualification

Research

Education Background

Query: "What is their highest degree and field of study?"
Use Case: Academic qualification verification

Industry Expertise

Query: "How many years have they worked in healthcare?"
Use Case: Industry-specific expertise validation

Geographic History

Query: "List all cities/countries where they've worked"
Use Case: Global experience assessment

Query Best Practices

Write Clear Queries

✅ Good: "How many years of software engineering experience?" ❌ Bad: "Experience?"

✅ Good: "Does this person mention AI or machine learning? Return Yes/No" ❌ Bad: "AI stuff"

Specify Output Format

For yes/no questions: "Return Yes/No"
For numbers: "Return as a number only"
For lists: "Return as comma-separated list"
For text: Specify max length if needed

Be Specific About Source

✅ Good: "What is their current job title at their current company?" ❌ Bad: "What's their title?" (ambiguous - current or previous?)

Processing Options

Lead Selection

All Leads

Processes every lead in the list
Best for comprehensive data coverage
Credits charged per lead processed

First 10 / First 20

Quick sampling of results
Test your query before processing all leads
Useful for large lists

Custom Count

Specify exactly how many leads to process
Process top N leads based on current sort
Flexible for any use case

Row-Level Indicators

Only selected leads show processing status
Other leads show "-" (not processed)
Easy to identify which leads were analyzed

Pricing

Smart Columns use AI processing credits:

Cost varies based on query complexity
Typically 1-5 credits per lead processed
Failed processing doesn't consume credits
View credit cost estimate before starting

Technical Details

AI Model

Uses advanced language models trained on business profiles
Context-aware understanding of professional information
Handles abbreviations, industry jargon, and implicit information
Supports multiple languages (profiles in non-English)

Data Sources

Smart Columns analyze:

LinkedIn profile summary
Work experience descriptions
Skills and endorsements
Education history
Certifications and licenses
Activity and posts (if available)

Processing Speed

Average: 3-8 seconds per lead
Parallel processing of multiple leads
Real-time progress updates
Results saved immediately upon completion

Accuracy

High confidence queries: 90-95% accuracy (e.g., "Current job title")
Inference required: 80-90% accuracy (e.g., "Budget authority")
Complex analysis: 70-85% accuracy (e.g., "Leadership style")

Limitations

What Smart Columns Can Do

✅ Extract information present in LinkedIn profiles ✅ Make reasonable inferences from profile content ✅ Analyze work history and experience ✅ Identify skills and technologies mentioned ✅ Determine seniority and role types

What Smart Columns Cannot Do

❌ Access private LinkedIn information ❌ Fetch real-time data not in the profile ❌ Predict future behavior or outcomes ❌ Access information behind LinkedIn premium features ❌ Violate LinkedIn's terms of service

Sorting and Filtering

Sorting Smart Columns

Click column header to sort by processed values
Alphabetical or numerical sorting (based on data type)
Empty values (-) sorted to bottom
Works with TanStack Table sorting

Filtering

Filter by specific values
Combine with other column filters
Export filtered results

Managing Smart Columns

Edit a Smart Column

Click the column settings icon
Modify the query definition
Re-process selected leads with new query

Delete a Smart Column

Remove column from the list
Processed data is permanently deleted
Credits are not refunded for deleted columns

Export Smart Column Data

Include smart columns in CSV exports
Download processed results
Integrate with CRM systems

Examples

Example 1: Sales Qualification

Goal: Identify VP-level prospects with budget authority

Column 1: "Is this person VP-level or higher? Return Yes/No"
Column 2: "Does this person mention budget or purchasing responsibility? Return Yes/No"

Result: Filter for Yes/Yes to get qualified leads

Example 2: Tech Stack Matching

Goal: Find prospects using Salesforce

Column: "Does this person mention using Salesforce in their experience or skills? Return Yes/No and include which role if yes"

Result: Targeted outreach for Salesforce-related products

Example 3: Hiring Pipeline

Goal: Find senior Python developers in fintech

Column 1: "How many years of Python development experience?"
Column 2: "How many years working in financial services or fintech?"

Result: Sort by experience to prioritize candidates

Troubleshooting

Column Shows "Failed"

LinkedIn profile may be incomplete or private
Query may be too complex or ambiguous
Try simplifying the query
No credits charged for failed processing

Unexpected Results

Review the query for clarity
Check if information exists in the LinkedIn profile
Consider making query more specific
Test on a few leads before processing all

Slow Processing

Complex queries take longer to process
Large batch sizes queue behind other jobs
Processing speed depends on LinkedIn data availability

Learn More

Enrichment Providers CRM Integrations