diff --git a/wiki/releasing-versions.md b/wiki/releasing-versions.md
index 0db48144ff8..aab3e1908a3 100644
--- a/wiki/releasing-versions.md
+++ b/wiki/releasing-versions.md
@@ -28,6 +28,15 @@ The latest changes have now been pushed to GitHub, a new `git` tag now exists on
 
 <sup>_\* GitHub Pages sites are cached aggressively and can sometimes take a couple of minutes to update._</sup>
 
+### Deploying to eui.elastic.co
+
+In addition to the GitHub pages deployment, we need to manually deploy to the EUI Bekitzur environment (which will eventually be the canonical home of the EUI docs).
+
+* Go to [the Jenkins job page](https://kibana-ci.elastic.co/job/elastic+eui+deploy-docs/build)
+* Log in (uses Elastic SSO)
+* Change the `branch_specifier` parameter to the new version tag name (e.g., `v48.0.0`)
+* Click "Build"
+
 ### Tag the release in GitHub
 
 We also update the [release's tag in github](https://github.com/elastic/eui/tags) by _creating a release_ for the version and copying over its _CHANGELOG.md_ entries. (TODO: screencast this next time to include a GIF here)
diff --git a/wiki/upgrading-kibana.md b/wiki/upgrading-kibana.md
index cbfd0c86031..8f05254ddc1 100644
--- a/wiki/upgrading-kibana.md
+++ b/wiki/upgrading-kibana.md
@@ -16,7 +16,8 @@ If the upgrade includes a large number of commits or has known breaking changes,
 	  * Or run `yarn test:jest` locally
 9. Resolve and commit reported issues (see [Resolving errors](#resolving-errors))
 10. Open a PR or mark existing as "Ready for review" (see [PR template](#pr-template))
-    * Add the `EUI`, `auto-backport`, `release_note:skip`, and relevant version tags
+    * Add the `EUI`, `ci:deploy-cloud`, `release_note:skip`, and relevant version tag(s) (the next minor version that has not yet entered feature freeze)
+	* We typically only merge into the main branch (the next minor version), but if an upgrade is specifically targeting a previous version, you can use the `auto-backport` tag to have Kibana automatically create a backport PR to an existing version branch.
     * Mention/ping any teams that are waiting on features to be available
 
 ## Resolving errors