--- a/testing/web-platform/tests/.azure-pipelines.yml
+++ b/testing/web-platform/tests/.azure-pipelines.yml
@@ -158,21 +158,24 @@ jobs:
     displayName: 'Run tests (Edge)'
   - task: PublishBuildArtifacts@1
     displayName: 'Publish results'
     inputs:
       artifactName: 'infrastructure'
     condition: always()
   - template: tools/ci/azure/cleanup_win10.yml
 
+# All `./wpt run` tests are run from epochs/* branches on a schedule. See
+# documentation at the top of this file for required setup.
 - job: results_edge
   displayName: 'all tests (Edge)'
-  # This job is only triggered manually until it has been shown to be robust.
-  condition: and(eq(variables['Build.Reason'], 'Manual'), variables['run_all_edge'])
-  # There are 5 agents in the pool, but use more jobs so that each takes <1h.
+  condition: |
+    or(eq(variables['Build.Reason'], 'Schedule'),
+       and(eq(variables['Build.Reason'], 'Manual'), variables['run_all_edge']))
+  # There are 12 agents in the pool, but use more jobs so that each takes <1h.
   strategy:
     parallel: 20
   timeoutInMinutes: 360
   pool:
     name: 'Hosted Windows Client'
   steps:
   - template: tools/ci/azure/system_info.yml
   - template: tools/ci/azure/checkout.yml
@@ -190,18 +193,16 @@ jobs:
     inputs:
       artifactName: 'edge-results'
   - template: tools/ci/azure/cleanup_win10.yml
 - template: tools/ci/azure/fyi_hook.yml
   parameters:
     dependsOn: results_edge
     artifactName: edge-results
 
-# All `./wpt run` tests are run from epochs/* branches on a schedule. See
-# documentation at the top of this file for required setup.
 - job: results_safari_preview
   displayName: 'all tests (Safari Technology Preview)'
   condition: |
     or(eq(variables['Build.Reason'], 'Schedule'),
        and(eq(variables['Build.Reason'], 'Manual'), variables['run_all_safari_preview']))
   strategy:
     parallel: 4 # chosen to make runtime ~2h
   timeoutInMinutes: 360