An important attribute of compound indexes is that they can be used for multiple queries on prefixes of the fields indexed. This is useful when we want to consolidate indexes that pile up in our collections over time.

Consider the compound (multi-field) index that we created previously:

> db.books.createIndex({"name": 1, "isbn": 1})

This can be used for queries on name or {name, isbn}:

> db.books.find({"name":"MongoDB Indexing"})
> db.books.find({"isbn": "1001", "name":"MongoDB Indexing"})

The order of the fields in our query doesn't matter; MongoDB will rearrange the fields to match our query.

However, the order of the fields in our index does matter. A query just for the isbn field cannot use our index:

> db.books.find({"isbn": "1001"})

The underlying reason is that our fields' values are stored in the index as secondary, tertiary, and so on; each one is embedded inside of the previous ones, just like a matryoshka, the Russian nesting doll. This means that when we query on the first field of our multi-field index, we can use the outermost doll to find our pattern, whereas when we are searching for the first two fields, we can match the pattern on the outermost doll, and then dive into the inner one.

This concept is called prefix indexing, and along with index intersection, it is the most powerful tool for index consolidation, as you will see later in this chapter.