diff --git a/docs/modules/clusters/pages/user-code-namespaces.adoc b/docs/modules/clusters/pages/user-code-namespaces.adoc
index 264fa8b2e..6ef0cd4b6 100644
--- a/docs/modules/clusters/pages/user-code-namespaces.adoc
+++ b/docs/modules/clusters/pages/user-code-namespaces.adoc
@@ -11,6 +11,7 @@ Your resources can be run on a Hazelcast cluster to support your business logic.
 
 * MapStore
 * Split brain events
+* Custom merge policies
 * `ExecutorService` runnable
 * Jet jobs
 
diff --git a/docs/modules/network-partitioning/pages/split-brain-recovery.adoc b/docs/modules/network-partitioning/pages/split-brain-recovery.adoc
index bc72c7afd..b9bb63878 100644
--- a/docs/modules/network-partitioning/pages/split-brain-recovery.adoc
+++ b/docs/modules/network-partitioning/pages/split-brain-recovery.adoc
@@ -254,6 +254,9 @@ public interface SplitBrainMergePolicy<V, T extends MergingValue<V>, R>
 NOTE: Please have in mind that `existingValue` can be `null`.
 This happens when a data structure or key-based entry was just created in the smaller cluster.
 
+NOTE: Custom merge policy classes must be present on the classpath of all member nodes. This can be achieved
+by updating the classpath for each node manually, or by using the {ucn} feature. See xref:clusters:user-code-namespaces.adoc[User Code Namespaces] for more information.
+
 === Merge Types
 
 A merge type defines an attribute which is required by a merge policy and provided by a data structure.