MongoDB Aggregation $out (original) (raw)

Last Updated : 16 Apr, 2026

The $out stage in MongoDB allows you to persist the results of an aggregation pipeline by writing them to a collection, making pre-processed data available for reporting, analytics, and future queries.

• $out is an aggregation pipeline stage that writes results to a specified collection.
• The output can be written to a new collection or replace an existing one.
• Must be the last stage in the aggregation pipeline.
• Stores results for later use without re-running complex aggregations.

Key Features of $out

The $out stage stores the final output of an aggregation pipeline into a collection, making processed data reusable for analytics, reporting, and future queries.

• **Stores Aggregation Results: Writes the pipeline output into a target collection for reuse.
• **Creates or Replaces Collections: Automatically creates the collection or overwrites existing data.
• **Optimizes Repeated Queries: Avoids re-running expensive pipelines by persisting results.
• **Supports Analytical Workloads: Helps build reporting and data-warehouse-style collections.
• **Final Stage Requirement: Must be the last stage in the aggregation pipeline.

**Syntax

{ $out: { db: "", coll: "" } }

• ****$out:** Specifies the stage in the aggregation pipeline.
• **db: Defines the target database for output.
• **coll: Specifies the target collection where results are stored

Example of Using $out in MongoDB

Below is a step-by-step guide demonstrating how to use the $out stage in an aggregation pipeline.

Step 1: Aggregating Data and Writing to a New Collection

Using the sample_supplies database, the pipeline groups documents by storeLocation, calculates total items sold per location, and writes the results to the store_items_totals collection.

Step1: SHOWING THE AGGREGATION PIPELINE

**Step 2: Verifying the Collection Creation

After running the query, a new collection is created, which can be verified using show collections; the store_items_totals collection appears in the sample_supplies database.

Step 2: LIST OF COLLECTIONS IN THE DATABASE

**Step 3:Checking Data in the New Collection

To check whether the query is written successfully in the new collection we run the command find( ) to display the items in the collections.

Step 3: DISPLAYING THE STORE_ITEMS_TOTALS COLLECTION

Step 4: Understanding the Output

Finally we can infer from the above image that our query is written successfully on a new collection. The results show aggregated sales data per store location.

Use Cases of MongoDB Aggregation $out

The $out stage is employed in various scenarios, including:

• **Data Warehousing: Storing pre-aggregated data in separate collections for analytical use.
• **Performance Optimization: Creating compact summary collections from large datasets for faster queries.
• **Precomputed Analytics: Persisting aggregated results to avoid repeated heavy computations.
• **Historical Snapshots: Maintaining aggregated collections to preserve historical trends and reduce operational data size.
• **Periodic Reporting: Saving scheduled aggregation results in dedicated collections for quick report generation.

outVsout Vs outVsmerge

MongoDB provides both outandout and outandmerge stages for storing aggregation results, but they serve different purposes.

Best Practices for Using $out

Here are some best practices of using $out:

• **Ensure the target collection is not capped: $out does not work with capped collections.
• **Avoid using outinshardedcollections:∗∗MongoDBdoesnotsupportout in sharded collections: MongoDB does not support outinshardedcollections:∗∗MongoDBdoesnotsupportout for sharded output collections.
• **Use indexes for efficiency: Index the newly created collection for faster queries.
• **Monitor Storage Impact: Since $out overwrites existing data, ensure you do not accidentally delete important information.
• **Consider mergeforincrementalupdates:∗∗Ifweneedtomergedatainsteadofoverwritingit,usemerge for incremental updates: If we need to merge data instead of overwriting it, use mergeforincrementalupdates:∗∗Ifweneedtomergedatainsteadofoverwritingit,usemerge.