LearnAI ToolsCareerPractice BuildsPlayContact
Lesson 815 min read

Documents & Collections

Understand the structure of a MongoDB document, the _id field, and how documents group into collections.

Anatomy of a Document

A document is a set of field-value pairs, similar to a JSON object. Values can be strings, numbers, booleans, dates, arrays, nested documents (embedded objects), or null.

{
  "_id": ObjectId("64f1a2b3c4d5e6f7a8b9c0d1"),
  "title": "Getting Started with MongoDB",
  "published": true,
  "views": 1420,
  "tags": ["database", "nosql", "mongodb"],
  "author": {
    "name": "Ada Lovelace",
    "email": "ada@example.com"
  },
  "createdAt": ISODate("2026-01-15T10:00:00Z")
}

The _id Field

Every document requires a unique _id field, serving as its primary key. If you don't provide one when inserting, MongoDB automatically generates an ObjectId — a 12-byte identifier that encodes a timestamp, making documents naturally sortable by creation time.

ObjectId Encodes a Timestamp

Because the first four bytes of an ObjectId represent its creation time, sorting documents by _id also roughly sorts them by insertion order — a useful, often-overlooked property.

Collections

A collection groups related documents together — roughly analogous to a table, but without a fixed schema enforced by default. A typical application database might have collections named users, posts, and orders.

use blog_app;
db.posts.insertOne({ title: "My First Post" }); // creates the "posts" collection if it doesn't exist

Documents Can Vary in Shape

Unlike a SQL table, two documents in the same collection are not required to share the same fields — a deliberate flexibility that's powerful, but requires discipline in application code to handle gracefully.

// Both valid documents in the same "products" collection
{ "_id": 1, "name": "T-Shirt", "size": "M", "color": "blue" }
{ "_id": 2, "name": "Laptop", "cpu": "M3", "ram": "16GB" }

Common Beginner Mistakes

Manually assigning a non-unique value to _id

MongoDB enforces uniqueness on _id automatically — attempting to insert a duplicate throws an error, which is exactly the intended safeguard.

Treating "flexible schema" as "no schema at all"

Application code still needs consistent expectations about document shape — the schema design lessons cover disciplined, deliberate approaches to this.

FAQs

Yes — any unique value (a string, a number) works as _id, though ObjectId's built-in uniqueness and timestamp encoding make it the sensible default for most cases.

Yes — 16MB per document, a limit that reinforces MongoDB's general design philosophy of documents representing one cohesive, not-too-large unit of data.

Summary

Documents are flexible, self-contained JSON-like records identified by a unique _id, grouped into collections. Next, you'll look closer at BSON and the specific data types MongoDB supports.

Next Lesson →

BSON & Data Types