Posted in

M365 Archive Gets File-Level Cold Storage — Cutting SharePoint Costs for MSP Clients

Stale files are quietly degrading your clients’ Copilot experience. Every outdated draft, obsolete project document, and abandoned report sitting in an active SharePoint site feeds into Copilot’s index. When a user asks for a summary of current project status, Copilot pulls from whatever it can see—including the three-year-old version of the spec that nobody remembered to archive. The result isn’t just bad search. It’s AI that confidently serves up wrong answers because the underlying content is stale.

Microsoft just shipped a fix for this, and it doubles as a storage cost win. File-level archiving in Microsoft 365 Archive is now in public preview. Instead of archiving entire SharePoint sites—which nobody does, because it kills access to the active stuff mixed in with the dead—you can now surgically move individual files and folders to cold storage. The active site keeps working. The stale files stop bloating the tenant and stop polluting Copilot.

What file-level archiving actually does

Until now, Microsoft 365 Archive worked at the site level. Move a whole site to cold storage, save money, done. But most SharePoint sites are a mix of active and stale—the HR site has current policies alongside onboarding documents from 2019. Site-level archiving was a non-starter for most tenants because it demanded an all-or-nothing decision.

File-level archiving changes that. An admin enables it per site, and then any user with edit permissions can archive individual files or folders. The archived files move to a lower-cost cold storage tier while retaining their permissions, retention policies, sensitivity labels, and legal hold settings. Everything stays inside the Microsoft 365 trust boundary. Reactivation is free—you pull a file back when you need it, no penalty.

The storage math is straightforward: Microsoft says you save up to 75% compared to buying additional SharePoint storage quota. Pay-as-you-go pricing only kicks in when the tenant exceeds its included allocation, so smaller tenants may archive without spending anything extra.

Syskit Point makes it scalable

Identifying which files to archive across a tenant with thousands of sites is the real problem. You cannot hand-audit this. Syskit Point provides the identification and governance layer—it uses both last-accessed and last-modified dates to find genuinely inactive files, not just old ones that people still reference. It can also trim excessive version history (SharePoint keeps up to 500 versions by default) before archiving, compounding the storage savings.

Microsoft’s own product lead, Brad Gussin, endorsed the integration directly in the announcement post. Syskit Point manages the full lifecycle from a single Storage Metrics report: identify, archive, restore, or delete. It is the governance layer, not the storage engine. The archiving itself is native M365 Archive.

What MSPs should do this week

This is a recurring-service opportunity that does not require a big upfront investment. Here is the play:

  • Pick one client tenant—ideally one approaching its SharePoint storage limit—and run a storage audit. Identify sites with the largest gap between total size and recently modified content.
  • Evaluate Syskit Point for that tenant. The governance layer is the prerequisite for doing file-level archiving at any scale beyond one-off manual cleanup.
  • Propose a managed archiving engagement: initial audit and policy setup, then monthly governance reviews. Frame it around both cost control and Copilot readiness—clients rolling out Copilot need clean data more than they need cheap storage.
  • Test the restore workflow. File-level reactivation is free, but make sure your team knows the process and can commit to an SLA before promising clients instant retrieval.

File-level archiving turns SharePoint storage management from a blunt quota-enforcement conversation into a value-add service. Your clients save money, their Copilot gets smarter, and you get a recurring engagement that does not depend on break-fix tickets.

Sources