# Best Practices

## Recommended setup

<table data-card-size="large" data-view="cards"><thead><tr><th></th><th></th></tr></thead><tbody><tr><td><strong>Use Custom SQL for filters and aggregations</strong></td><td>Rather than exporting a full table and filtering in your destination, write a Custom SQL query in Coupler.io. This reduces data transfer, keeps your destination lean, and makes your data flow self-documenting.</td></tr><tr><td><strong>Test queries in Redshift first</strong></td><td>Before pasting a Custom SQL query into Coupler.io, run it directly in your Redshift query editor (e.g., Redshift Query Editor v2). This lets you verify the results and execution time without tying up the data flow.</td></tr><tr><td><strong>Use lowercase schema and table names</strong></td><td>Redshift treats unquoted identifiers as lowercase. To avoid case-sensitivity issues, use lowercase names in your Custom SQL or quote them: `"MyTable"` for mixed case.</td></tr></tbody></table>

## Data refresh and scheduling

<table data-card-size="large" data-view="cards"><thead><tr><th></th><th></th></tr></thead><tbody><tr><td><strong>Schedule after your data load window</strong></td><td>If your Redshift cluster processes data at 2 AM, schedule your Coupler.io refresh for 2:30 AM (your timezone) to ensure fresh data. Add a buffer for large ETL jobs.</td></tr><tr><td><strong>Run a manual test first</strong></td><td>Always execute a manual run of your data flow before scheduling. This catches connection issues, query errors, and unexpected data volumes before automation kicks in.</td></tr></tbody></table>

## Performance optimization

<table data-card-size="large" data-view="cards"><thead><tr><th></th><th></th></tr></thead><tbody><tr><td><strong>Avoid SELECT *; name your columns</strong></td><td>Explicitly list the columns you need: `SELECT id, name, email FROM customers`. This is faster, makes your query clearer, and reduces accidental data leaks if new columns are added to the table.</td></tr><tr><td><strong>Filter early in the query</strong></td><td>Use WHERE clauses to reduce rows before aggregating or joining: `SELECT ... FROM orders WHERE order_date >= CURRENT_DATE - 30 DAYS`. Filtering in Redshift is faster than filtering in your destination.</td></tr><tr><td><strong>Use SORTKEY and DISTKEY for large tables</strong></td><td>If you're querying the same large table repeatedly, ask your Redshift administrator to set SORTKEY and DISTKEY columns that match your filter and join patterns. This can dramatically speed up queries.</td></tr></tbody></table>

## Common pitfalls

{% columns %}
{% column %}
**Do**

* Test your data flow with a manual run before scheduling
* Write readable Custom SQL with comments for complex logic
* Monitor the first few scheduled runs to confirm data arrives correctly
* Keep destination sheets or tables organized with clear column headers
  {% endcolumn %}

{% column %}
**Don't**

* Export raw tables without filtering — plan your queries in advance
* Use hardcoded dates in WHERE clauses; instead, use `CURRENT_DATE` or `DATEADD()` for dynamic ranges
* Run multiple large data flows simultaneously on a single Redshift cluster during peak hours
* Store sensitive credentials (passwords, API keys) in your Custom SQL queries — use Coupler.io's connection settings
  {% endcolumn %}
  {% endcolumns %}

{% hint style="danger" %}
**Avoid temporary tables:** If you use `CREATE TEMP TABLE` in a Custom SQL query, the table won't persist between Coupler.io's connection checks and data fetch. Use permanent tables or views instead.
{% endhint %}