diff --git a/mug/src/main/java/com/google/mu/collect/Sequence.java b/mug/src/main/java/com/google/mu/collect/Sequence.java
index 61ffa9511e..813e4c7127 100644
--- a/mug/src/main/java/com/google/mu/collect/Sequence.java
+++ b/mug/src/main/java/com/google/mu/collect/Sequence.java
@@ -32,10 +32,12 @@
  * recursive. That is, if your Sequence is the result of 1 million concatenations, you won't run
  * into stack overflow error because under the hood, it's a heap-allocated immutable tree structure.
  *
- * <p>The expected use case is to perform frequent concatenations using the {@code concat()}
- * methods. O(n) materialization cost will be (lazily) paid before the first time accessing the
- * elements through the {@link List} interface such as {@link List#get}, {@link List#equals},
- * or {@link #toString} etc.
+ * <p>The expected use case is to concatenate lots of smaller {@code Sequence}s using the {@code
+ * concat()} methods to create the final Sequence. O(n) materialization cost will be (lazily) paid
+ * before the first time accessing the elements of the final Sequence through the {@link List}
+ * interface such as {@link List#get}, {@link List#equals}, {@link #toString} etc. You may also
+ * want to copy the final Sequence into a more conventional List such as {@link
+ * com.google.common.collect.ImmutableList#copyOf Guava ImmutableList}.
  *
  * <p>On the other hand, it's inefficient to materialize, concatenate then materialize the
  * concatenated Sequence...