diff --git a/code/services-core/control-service/src/main/resources/templates/control/sys/new-domain-ranking-set.hdb b/code/services-core/control-service/src/main/resources/templates/control/sys/new-domain-ranking-set.hdb
index a4fa720e..2907a824 100644
--- a/code/services-core/control-service/src/main/resources/templates/control/sys/new-domain-ranking-set.hdb
+++ b/code/services-core/control-service/src/main/resources/templates/control/sys/new-domain-ranking-set.hdb
@@ -44,8 +44,12 @@
A list of domain names, one per line, possibly globbed with SQL-style '%' wildcards.
- These are used as the origin point for the Personalized PageRank algorithm, and will be considered
- the central points of the link or adjacency graph. If no domains are specified, the entire domain space is used, as per the PageRank paper.
+ If provided, these are used as the origin point for the Personalized PageRank algorithm, and will be considered
+ the central points of the link or adjacency graph. If no domains are specified, the entire domain space is used,
+ as per the PageRank paper.
+
+ If similarity data is available and domains are specified, the similarity data is used as basis for the ranking
+ calculation instead, providing a much more coherent ranking.
diff --git a/code/services-core/control-service/src/main/resources/templates/control/sys/update-domain-ranking-set.hdb b/code/services-core/control-service/src/main/resources/templates/control/sys/update-domain-ranking-set.hdb
index d906e871..ffd9a12b 100644
--- a/code/services-core/control-service/src/main/resources/templates/control/sys/update-domain-ranking-set.hdb
+++ b/code/services-core/control-service/src/main/resources/templates/control/sys/update-domain-ranking-set.hdb
@@ -46,9 +46,12 @@
A list of domain names, one per line, possibly globbed with SQL-style '%' wildcards.
- These are used as the origin point for the Personalized PageRank algorithm, and will be considered
- the central points of the link or adjacency graph. If no domains are specified,
- the entire domain space is used, as per the PageRank paper.
+ If provided, these are used as the origin point for the Personalized PageRank algorithm, and will be considered
+ the central points of the link or adjacency graph. If no domains are specified, the entire domain space is used,
+ as per the PageRank paper.
+
+ If similarity data is available and domains are specified, the similarity data is used as basis for the ranking
+ calculation instead, providing a much more coherent ranking.