diff --git a/CHANGES.md b/CHANGES.md
index e02dc0dca..4cfce1219 100644
--- a/CHANGES.md
+++ b/CHANGES.md
@@ -5,8 +5,10 @@
 ### Bug fixes
 
 * filter, merge: Fixed formatting of the error message shown when there are duplicate sequence ids. [#1954][] @victorlin
+* filter: Adjusted the error message shown when there are missing weights to mention the option of updating values in metadata. [#1956][] @victorlin
 
 [#1954]: https://github.com/nextstrain/augur/pull/1954
+[#1956]: https://github.com/nextstrain/augur/pull/1956
 
 ## 33.0.0 (26 January 2026)
 
diff --git a/augur/filter/subsample.py b/augur/filter/subsample.py
index 7ca6b1e97..e372d133b 100644
--- a/augur/filter/subsample.py
+++ b/augur/filter/subsample.py
@@ -486,9 +486,10 @@ def _handle_incomplete_weights(
         raise AugurError(dedent(f"""\
             The input metadata contains these values under the following columns that are not covered by {weights_file!r}:
             - {columns_with_values}
-            To fix this, either:
-            (1) specify weights explicitly - add entries to {weights_file!r} for the values above, or
-            (2) specify a default weight - add an entry to {weights_file!r} with the value {COLUMN_VALUE_FOR_DEFAULT_WEIGHT!r} for all columns"""))
+            Some ways to fix this:
+            (1) use existing weights - update the metadata file to use values in {weights_file!r}, or
+            (2) specify new weights - add entries to {weights_file!r} for the values above, or
+            (3) specify a default weight - add an entry to {weights_file!r} with the value {COLUMN_VALUE_FOR_DEFAULT_WEIGHT!r} for all columns"""))
     else:
         print_err(dedent(f"""\
             WARNING: The input metadata contains these values under the following columns that are not directly covered by {weights_file!r}:
diff --git a/tests/functional/filter/cram/subsample-weighted.t b/tests/functional/filter/cram/subsample-weighted.t
index b34680af7..626c13dcc 100644
--- a/tests/functional/filter/cram/subsample-weighted.t
+++ b/tests/functional/filter/cram/subsample-weighted.t
@@ -89,9 +89,10 @@ Sampling with incomplete weights should show an error.
   Sampling with weights defined by weights.tsv.
   ERROR: The input metadata contains these values under the following columns that are not covered by 'weights.tsv':
   - 'location': ['B']
-  To fix this, either:
-  (1) specify weights explicitly - add entries to 'weights.tsv' for the values above, or
-  (2) specify a default weight - add an entry to 'weights.tsv' with the value 'default' for all columns
+  Some ways to fix this:
+  (1) use existing weights - update the metadata file to use values in 'weights.tsv', or
+  (2) specify new weights - add entries to 'weights.tsv' for the values above, or
+  (3) specify a default weight - add an entry to 'weights.tsv' with the value 'default' for all columns
   [2]
 
 Re-running with a default weight shows a warning and continues.
@@ -139,9 +140,10 @@ To specify a default weight, the value 'default' must be set for all weighted co
   ERROR: The input metadata contains these values under the following columns that are not covered by 'weights.tsv':
   - 'location': ['B']
   - 'month': ['2000-01', '2000-03']
-  To fix this, either:
-  (1) specify weights explicitly - add entries to 'weights.tsv' for the values above, or
-  (2) specify a default weight - add an entry to 'weights.tsv' with the value 'default' for all columns
+  Some ways to fix this:
+  (1) use existing weights - update the metadata file to use values in 'weights.tsv', or
+  (2) specify new weights - add entries to 'weights.tsv' for the values above, or
+  (3) specify a default weight - add an entry to 'weights.tsv' with the value 'default' for all columns
   [2]
 
   $ cat >weights.tsv <<~~