diff --git a/meeting-notes/20190311.md b/meeting-notes/20190311.md
index a3da78493425be984e88156ae09a0ce08885868d..2ab49c7396e3407d0622231d72de7586909a99e8 100644
--- a/meeting-notes/20190311.md
+++ b/meeting-notes/20190311.md
@@ -1,9 +1,5 @@
 # Meeting 11.03.2019
 
-## Questions
-
-* Collect further data? IVs? /Stuart?
-
 ## Discussion/Announcements
 
 * working to bring back abuse_filter_history table
@@ -13,6 +9,9 @@
 
 ## TODO
 
+* Frame in context: how do Wikipedia quality control mechanisms work? What is the contribution of edit filters?
+* Why do filters exist? Why don't they just use Huggle?
+* Hypothesis: filter should be used for frequently occurring cases (there's also the note that every edit is checked against all enabled filters, so it's not worth it to create filters for phenomena that occur rarely)? What issues are apparently frequently occurring?
 
 ## Fun Facts