diff --git a/keps/sig-apps/2307-job-tracking-wihout-lingering-pods/README.md b/keps/sig-apps/2307-job-tracking-wihout-lingering-pods/README.md
index 60693c0e7874..e8330e4101ef 100644
--- a/keps/sig-apps/2307-job-tracking-wihout-lingering-pods/README.md
+++ b/keps/sig-apps/2307-job-tracking-wihout-lingering-pods/README.md
@@ -223,6 +223,9 @@ type UncountedTerminatedPods struct {
 }
 ```
 
+Note: the final name of the field `uncountedTerminatedPods` will be decided
+during API review.
+
 ### Algorithm
 
 The following algorithm updates the status counters without relying on finished
@@ -281,9 +284,9 @@ This is different from the legacy tracking, where the Job controller does not
 account for deleted Pods. This is a limitation that this KEP also wants to
 solve.
 
-However, if it's the Job controller what deletes the Pod (when parallelism
-is decreased, for example), the controller removes the finalizer before
-deleting it. Thus, these deletions don't count towards the failures.
+However, if the Job controller deletes the Pod (when parallelism is decreased,
+for example), the controller removes the finalizer before deleting it. Thus,
+these deletions don't count towards the failures.
    
 ### Pod adoption