Run a Query

Run a Query executes SQL queries against your BigQuery datasets to retrieve, analyze, and transform data for your marketing workflows.

​

When to Use It

• Execute custom SQL queries to retrieve specific data
• Generate reports and analytics for marketing campaigns
• Use it as AI Tool for AI Agents to run dynamic SQL queries

​

Inputs

​

Outputs

​

Credit Cost

1 credit per operation.

​

Real-World Examples

BigQuery AI Agent:

AI Agent → List Datasets → List Tables → Get Table Schema → Run Query
"Allow AI to fully explore BigQuery and run intelligent SQL queries"

Monthly Campaign Analysis:

Run Query → AI Analyze Data → Send Email
"Execute custom SQL to analyze campaign performance and email insights"

​

Understanding Query Results

The returned data includes:

Query Output:

• Array of rows with your query results
• Column names matching your SELECT statement
• Data types preserved from BigQuery tables
• Null values handled appropriately

​

SQL Query Examples

Campaign Performance Analysis:

SELECT 
  campaign_name,
  SUM(spend) as total_spend,
  SUM(conversions) as total_conversions,
  ROUND(SUM(conversions) / SUM(spend), 2) as cost_per_conversion
FROM marketing_campaigns 
WHERE DATE(created_at) >= DATE_SUB(CURRENT_DATE(), INTERVAL 30 DAY)
GROUP BY campaign_name
ORDER BY total_conversions DESC

​

Best Practices

Query Optimization:

• Use WHERE clauses to limit data scanning and reduce costs
• Select only the columns you need rather than using SELECT *
• Filter by date ranges to avoid processing unnecessary historical data
• Use LIMIT for testing queries before running on full datasets

Cost Management:

• Preview query costs in BigQuery console before execution
• Use partitioned tables and filter by partition keys
• Monitor bytes processed to manage BigQuery costs
• Consider creating views for frequently used complex queries

Performance:

• Use proper data types and avoid unnecessary conversions
• Leverage clustering and partitioning for large tables
• Test complex queries with LIMIT before full execution
• Monitor execution times and optimize as needed

​

Tips

SQL Development:

• Test your queries in BigQuery console first
• Use query validation to check syntax before execution
• Comment your SQL for team documentation
• Save frequently used queries as templates

Dynamic Queries:

• Use parameterized queries for flexible filtering
• Combine with other workflow nodes for dynamic table/column names
• Consider using variables from previous workflow steps
• Plan for different data types and null value handling

Integration:

• Results integrate seamlessly with other Markifact nodes
• Use with AI Analyze Data for insights on query results
• Combine with Rename Fields to standardize output column names
• Perfect for feeding data into Google Sheets or other destinations

​

FAQ