diff --git a/website/docs/sql-reference/aggregate-functions/sql-array-agg.md b/website/docs/sql-reference/aggregate-functions/sql-array-agg.md
index 9f4af7ca1fc..a6f508a7bef 100644
--- a/website/docs/sql-reference/aggregate-functions/sql-array-agg.md
+++ b/website/docs/sql-reference/aggregate-functions/sql-array-agg.md
@@ -59,4 +59,4 @@ Looking at the query results—this makes sense! We’d expect newer orders to l
There are definitely too many use cases to list out for using the ARRAY_AGG function in your dbt models, but it’s very likely that ARRAY_AGG is used pretty downstream in your since you likely don’t want your data so bundled up earlier in your DAG to improve modularity and dryness. A few downstream use cases for ARRAY_AGG:
- In [`export_` models](https://www.getdbt.com/open-source-data-culture/reverse-etl-playbook) that are used to send data to platforms using a tool to pair down multiple rows into a single row. Some downstream platforms, for example, require certain values that we’d usually keep as separate rows to be one singular row per customer or user. ARRAY_AGG is handy to bring multiple column values together by a singular id, such as creating an array of all items a user has ever purchased and sending that array downstream to an email platform to create a custom email campaign.
-- Similar to export models, you may see ARRAY_AGG used in [mart tables](https://docs.getdbt.com/best-practices/how-we-structure/4-marts) to create final aggregate arrays per a singular dimension; performance concerns of ARRAY_AGG in these likely larger tables can potentially be bypassed with use of [incremental models in dbt](https://docs.getdbt.com/docs/build/incremental-models).
+- Similar to export models, you may see ARRAY_AGG used in [mart tables](/best-practices/how-we-structure/4-marts) to create final aggregate arrays per a singular dimension; performance concerns of ARRAY_AGG in these likely larger tables can potentially be bypassed with use of [incremental models in dbt](/docs/build/incremental-models).