MongoDB $group and $project stages explained with examples

$group — collapse documents into summaries

$group collapses all documents sharing the same _id expression value into a single output document. Accumulator operators compute values across all documents in each group. Set _id: null to aggregate across the entire collection into one result.

// Group products by category and compute stats per group
const stats = await db.collection('products').aggregate([
  { $group: {
    _id: '$category',
    productCount: { $sum: 1 },        // count documents
    avgPrice:     { $avg: '$price' }, // mean price
    minPrice:     { $min: '$price' }, // lowest price
    maxPrice:     { $max: '$price' }, // highest price
    allNames:     { $push: '$name' }, // array with duplicates allowed
    uniqueTags:   { $addToSet: '$tag' } // deduplicated array
  }}
]).toArray()

$project — reshape and compute fields

$project includes, excludes, renames, and computes new fields using aggregation expressions. Unlike a find projection, $project can create entirely new computed fields using operators like $multiply, $concat, and $year.

db.collection('orders').aggregate([
  { $project: {
    _id: 0,
    orderId: '$_id',
    discountedTotal: { $multiply: ['$total', 0.9] },
    customerName: { $concat: ['$firstName', ' ', '$lastName'] },
    orderYear: { $year: '$createdAt' },
    isPremium: { $gt: ['$total', 500] }
  }}
])

When working with $group, remember that the _id field defines the grouping key — it is required and must not be confused with the document's _id field from the collection. Setting _id: null aggregates all documents in the pipeline into a single output document, which is useful for computing collection-wide totals like overall revenue or mean order value across all records.