diff --git a/src/main/resources/hudson/plugins/ec2/AmazonEC2Cloud/help-noDelayProvisioning.html b/src/main/resources/hudson/plugins/ec2/AmazonEC2Cloud/help-noDelayProvisioning.html
index 00e0f0b66..9f9e44ee6 100644
--- a/src/main/resources/hudson/plugins/ec2/AmazonEC2Cloud/help-noDelayProvisioning.html
+++ b/src/main/resources/hudson/plugins/ec2/AmazonEC2Cloud/help-noDelayProvisioning.html
@@ -1,5 +1,9 @@
-By default Jenkins do estimate load to avoid over-provisioning of cloud nodes.
-With this option enabled, a new node is created on Amazon EC2 as soon as NodeProvisioner detects need for more agents.
-In worse scenarios, this will results in some extra nodes provisioned on EC2, which will be shortly terminated. EC2
-being billed by the second since 2017 this won't have a huge impact on infrastructure costs, while offering build nodes
-without unnecessary delays.
\ No newline at end of file
+Jenkins avoids over-provisioning of cloud nodes by examining the current load
+and deciding when more nodes are needed. Enable "No delay provisioning" to
+create a new cloud node on Amazon EC2 as soon as the NodeProvisioner detects a
+need for more agents. This may cause Jenkins to create more cloud nodes than
+it needs to serve the current jobs. These nodes will be terminated if they
+remain idle longer than the "idle termination time" setting. EC2 currently
+bills instances by the second, so this option should not have a large impact
+on your AWS bill. It will make cloud nodes available much more quickly for
+jobs and should reduce delays.
\ No newline at end of file