diff --git a/website/blog/2023-2-17-large-exchange-backups.md b/website/blog/2023-2-17-large-exchange-backups.md
new file mode 100644
index 000000000..9b7640104
--- /dev/null
+++ b/website/blog/2023-2-17-large-exchange-backups.md
@@ -0,0 +1,65 @@
+---
+slug: large-microsoft-365-exchange-backups
+title: "Backing up large Microsoft 365 Exchange mailboxes with Corso"
+description: "A guide to using Corso to back up very large Exchange mailboxes in Microsoft 365"
+authors: nica
+tags: [corso, microsoft 365, backups, S3]
+date: 2023-2-21
+image: ./images/heavy-mover.jpg
+---
+
+![heavy earth mover By Lowcarb23 - Own work, CC BY-SA 4.0, https://commons.wikimedia.org/w/index.php?curid=114344394](./images/heavy-mover.jpg)
+
+Over the last few months it’s been amazing sharing Corso with more and more users. One pleasant surprise has been users
+who are operating in large, often multi-tenant deployments of Microsoft 365 who want to use Corso to back up all their
+data. In our discussions on the [Corso User Discord](https://discord.gg/63DTTSnuhT), we’ve found some best practices for
+backing up large Exchange mailboxes with Corso.
+
+<!-- truncate -->
+
+### Make sure you’re using the latest version of Corso
+
+We've recently done a lot of work  to harden Corso against transient network outages and Graph API timeouts. This
+hardening work makes the most impact during large backups as their long runtime increase the probability of running
+into transient errors.
+
+Our recent work has also included support for incremental backups, which you’ll definitely need for larger data sets.
+This means that while your first backup of a user with a large mailbox can take some time, all subsequent backups
+will be quite fast as Corso will only capture the incremental changes while still constructing a full backup.
+
+### Don’t be afraid to restart your backups
+
+Fundamentally, Corso is a consumer of the Microsoft Graph API, which like all complex API’s, isn’t 100% predictable.
+Even in the event of a failed backup, Corso will often have stored multiple objects in the course of a backup. Corso
+will work hard to reuse these stored objects in the next backup, meaning your next backup isn’t starting from
+zero. A second attempt is likely to run faster with a better chance of completing successfully.
+
+### Batch your users
+
+If many of your users have large file attachments (or if you have more than a few hundred users), you’ll want to batch
+your users for your first backup. A tool like [Microsoft365dsc](https://microsoft365dsc.com/) can help you get a list
+of all user emails ready for parsing. After that you can back up a few users or even a single user at a time with the
+Corso command `corso backup create exchange --user "alice@example.com,bob@example.com"`
+
+Why can’t you just run them all in one go with `--user '*'` ? Again we’re limited by the Microsoft’s Graph API which
+often has timeouts, 5xx errors, and throttles its clients.
+
+The good news is that with Corso’s robust ability to do incremental backups, after your first backup, you can
+absolutely use larger batches of users, as all future backups to the same repository will run **much** faster.
+
+### Use multiple repositories for different tenants
+
+If you’re a managed service provider or otherwise running a multi-tennant architecture, you should use multiple separate
+repositories with Corso. Two ways to pursue this:
+
+- Point to separate buckets
+- Place other repositories in subfolders of the same bucket with the `prefix` option
+
+In both cases, the best way to keep these settings tidy is by using multiple `.corso.toml`
+[configuration files](../../docs/setup/configuration/#configuration-file). Use the
+`-config-file` option to point to separate config files
+
+### Have questions?
+
+Corso is under active development, and we expect our support for this type of use case to improve rapidly.
+If you have feedback for us please do [join our discord](https://discord.gg/63DTTSnuhT) and talk directly with the team!
diff --git a/website/blog/images/heavy-mover.jpg b/website/blog/images/heavy-mover.jpg
new file mode 100644
index 000000000..b2174fdaf
Binary files /dev/null and b/website/blog/images/heavy-mover.jpg differ