$facet, $bucket, and multi-pipeline aggregations

Why this matters

$facet lets you run multiple aggregation sub-pipelines over the same input documents in a single pass, which is how faceted search ("show me counts per category AND per price range AND per rating simultaneously") works without three separate queries.$bucket groups continuous numeric values into defined ranges — the building block for histograms, price range filters, and age distribution charts. Understanding these stages matters because the naive alternative is running three queries and merging results in application code, which is 3× the database load, 3× the latency, and produces inconsistent snapshots since each query runs at a different moment.

Demo

Compute category counts and price-range buckets in a single aggregation pass using $facet.

Try it yourself

1. Add a third facet named topByRating that uses $sort and $limit to return the top 3 highest-rated products. Confirm all three facets return consistent results from the same pipeline invocation.
2. Change the $bucket `boundaries` so one range spans only$1 (e.g. [49, 50]). Insert a product at exactly $49.99. Confirm which bucket it falls into and verify the boundary semantics (inclusive lower, exclusive upper).
3. Use $bucketAuto with buckets: 4 instead of manual boundaries. Compare the auto-generated boundaries to your manual ones. Note which approach produces more even distribution for your dataset.

Prompt your AI

Use these three in order. Each builds on the one before.

1. Basics & terminology

Explain MongoDB's $facet stage. What problem does it solve compared to running multiple separate aggregations?

2. Why it works (the mechanism)

When $facet runs two sub-pipelines, does MongoDB read the input documents once or twice? What are the memory implications for large input sets?

3. Advanced — application & what's next

I'm building a product search page that needs category counts, price range counts, and brand counts all computed at query time. Design a $facet pipeline that also accepts a user's filter (e.g. selected category) and correctly shows counts for the other facets as if that filter is applied.

References

• $facet — MongoDB docs
• $bucket — MongoDB docs