diff --git a/src/agenda_culturel/import_tasks/extractor.py b/src/agenda_culturel/import_tasks/extractor.py
index 4822dad..3e75870 100644
--- a/src/agenda_culturel/import_tasks/extractor.py
+++ b/src/agenda_culturel/import_tasks/extractor.py
@@ -2,7 +2,7 @@ from abc import ABC, abstractmethod
 from datetime import datetime, time, date, timedelta
 import re
 import unicodedata
-
+from django.utils import timezone
 
 
 
@@ -240,6 +240,28 @@ class Extractor(ABC):
         from .extractor_ggcal_link import GoogleCalendarLinkEventExtractor
 
         if single_event:
-            return [FacebookEventExtractor(), GoogleCalendarLinkEventExtractor()]
+            return [FacebookEventExtractor(), GoogleCalendarLinkEventExtractor(), EventNotFoundExtractor()]
         else:
-            return [ICALExtractor(), FacebookEventExtractor(), GoogleCalendarLinkEventExtractor()]
+            return [ICALExtractor(), FacebookEventExtractor(), GoogleCalendarLinkEventExtractor(), EventNotFoundExtractor()]
+
+
+# A class that only produce a not found event
+class EventNotFoundExtractor(Extractor):
+
+    def extract(
+        self, content, url, url_human=None, default_values=None, published=False
+    ):
+        self.set_header(url)
+        self.clear_events()
+
+        self.add_event(default_values, "événement sans titre", 
+            None, timezone.now().date(), None, 
+            "l'import a échoué, la saisie doit se faire manuellement à partir de l'url source",
+            [],  [url], published=False, url_human=url)
+
+        return self.get_structure()
+
+
+    def clean_url(url):
+        return url
+
diff --git a/src/agenda_culturel/import_tasks/generic_extractors.py b/src/agenda_culturel/import_tasks/generic_extractors.py
index 03061e1..10803c9 100644
--- a/src/agenda_culturel/import_tasks/generic_extractors.py
+++ b/src/agenda_culturel/import_tasks/generic_extractors.py
@@ -269,4 +269,5 @@ class TwoStepsExtractorNoPause(TwoStepsExtractor):
         result = super().extract(content, url, url_human, default_values, published, only_future, ignore_404)
         self.downloader.pause = pause
 
-        return result
\ No newline at end of file
+        return result
+
diff --git a/src/agenda_culturel/import_tasks/importer.py b/src/agenda_culturel/import_tasks/importer.py
index 753381f..f9c0b0a 100644
--- a/src/agenda_culturel/import_tasks/importer.py
+++ b/src/agenda_culturel/import_tasks/importer.py
@@ -1,6 +1,11 @@
 from .downloader import *
 from .extractor import *
 
+import logging
+
+logger = logging.getLogger(__name__)
+
+
 
 class URL2Events:
     def __init__(
@@ -29,8 +34,9 @@ class URL2Events:
         else:
             # if the extractor is not defined, use a list of default extractors
             for e in Extractor.get_default_extractors(self.single_event):
+                logger.warning('Extractor::' + type(e).__name__)
                 e.set_downloader(self.downloader)
                 events = e.extract(content, url, url_human, default_values, published)
-                if events is not None:
+                if events is not None and len(events) > 0:
                     return events
         return None
diff --git a/src/agenda_culturel/models.py b/src/agenda_culturel/models.py
index d4853a6..022dd6d 100644
--- a/src/agenda_culturel/models.py
+++ b/src/agenda_culturel/models.py
@@ -1490,7 +1490,7 @@ class Event(models.Model):
         imported = Event.objects.bulk_create(to_import)
         # update organisers (m2m relation)
         for i, ti in zip(imported, to_import):
-            if ti.has_pending_organisers():
+            if ti.has_pending_organisers() and ti.pending_organisers is not None:
                 i.organisers.set(ti.pending_organisers)
 
         nb_updated = Event.objects.bulk_update(