Exit costs and vendor lock-in in backup software
Leaving a backup vendor can cost more than the software itself: not usually as one visible invoice, but as a proprietary storage format only that vendor's tooling can read, an egress fee on cloud-hosted data, and the operational effort of re-onboarding every mailbox somewhere else. A backup product's real price includes what it costs to stop using it, not just what it costs to start.
What's actually documented, vendor by vendor
Two sourced examples of the "your data lives in our cloud" model: Veeam Data Cloud stores backups in Veeam-operated Azure storage accounts, with no customer-controlled storage option documented (sourced comparison). NinjaOne SaaS Backup (formerly Dropsuite) is hosted entirely on the vendor's own AWS infrastructure, with no self-hosted option at all (sourced comparison). It's still worth crediting directly that Dropsuite's own pages describe free EML/PST export, a genuinely good point on exit, even for a fully SaaS-hosted product. MailStore, by contrast, documents several supported export formats (EML/MSG, PST, or direct export into another mailbox) but does not openly specify its own internal storage format (sourced comparison). Export works, but you're relying on MailStore's own export tooling to get it, not an independently-readable format.
What "open format" needs to actually mean
Osiris's chunk-store format is open and documented, specifically so that a small standalone tool can restore data from it without a running Osiris server. Seestorage andYour data is your data. That standalone-restore capability is treated as a release requirement, not a someday feature: seehow Osiris is built for where it sits in the release checklist. The AGPL-3.0 licence adds a second layer to the same guarantee (seeopen source) since the code that reads the format is itself inspectable and forkable, not just the format's specification.
Frequently asked
What does it cost to switch backup vendors?
Three things, typically: re-pointing every mailbox at the new tool and letting it re-seed a first backup (time and, on a metered ingest, sometimes money), any egress or export fee the old vendor charges to get your data out, and whatever is not recoverable at all if the old vendor's storage format has no independent reader. The third one is the expensive one, because it isn't a fee: it's data you simply don't get back.
Does an open storage format alone solve vendor lock-in?
Mostly, but not completely: you still need a way to actually read that format without the vendor's own running server, or "open" is theoretical. That's why Osiris ships a standalone restore tool alongside the documented format, not just a specification: openness that requires the original software to exercise it isn't much of an exit.
Vendor facts sourced from Veeam's, Dropsuite's and MailStore's own official documentation, checked 2026-09-23/24. See the linked comparison pages for exact URLs and dates.