diff --git a/.gitlab-ci.yml b/.gitlab-ci.yml
index d025e6d449d5d5aeb0b97f1c3bf8d19a73d6bcc3..f0537e236fd4ce58cec537007b49471499a10858 100644
--- a/.gitlab-ci.yml
+++ b/.gitlab-ci.yml
@@ -220,6 +220,8 @@ stages:
   timeout: "240 minutes"
   variables:
     MLD_ARTIFACT_NAME: "mld--$CI_JOB_NAME-$CI_JOB_ID--sha-$CI_COMMIT_SHORT_SHA.csv"
+    MERGED_CSV_ARTIFACT_NAME: "$CI_JOB_NAME--merged_csv--$CI_JOB_ID.csv"
+    PAGES_HTML_ARTIFACT_NAME: "$CI_JOB_NAME-index.html"
   script:
     - *print-common-info
     - *update-scripts-repo
@@ -250,11 +252,11 @@ stages:
     -    unzip artifacts.zip -d previous_artifacts
     # This wildcard thingy relies on only one csv file being present per job
     -    file_previous="previous_artifacts/mld--ivas-pytest-mld-long-dec-$id_previous--sha-*.csv"
-    -    python3 ci/basop-pages/create_report_pages.py $CI_JOB_NAME-index.html $CI_JOB_NAME--merged_csv--$CI_JOB_ID.csv $MLD_ARTIFACT_NAME $file_previous $CI_JOB_ID $id_previous $CI_JOB_NAME
+    -    python3 ci/basop-pages/create_report_pages.py $PAGES_HTML_ARTIFACT_NAME $MERGED_CSV_ARTIFACT_NAME $MLD_ARTIFACT_NAME $file_previous $CI_JOB_ID $id_previous $CI_JOB_NAME
     - else
     # create empty file for artifacts to avoid errors
     -    touch $CI_JOB_NAME-index.html
-    -    touch $CI_JOB_NAME--merged_csv--$CI_JOB_ID.csv
+    -    touch $MERGED_CSV_ARTIFACT_NAME
     - fi
 
     - if [ $zero_errors != 1 ]; then echo "Run errors encountered!"; exit $EXIT_CODE_FAIL; fi
@@ -271,10 +273,9 @@ stages:
     paths:
       - report-junit.xml
       - report.html
-      - $CI_JOB_NAME-index.html
+      - $PAGES_HTML_ARTIFACT_NAME
       - $MLD_ARTIFACT_NAME
-      # the * wildcard is needed because $id_previous is undefined here
-      - $CI_JOB_NAME-merged_csv--$CI_JOB_ID.csv
+      - $$MERGED_CSV_ARTIFACT_NAME
     expose_as: "pytest mld results"
     reports:
       junit: