PRODUCT TOUR
How MigrationFox works, end to end
Every migration follows the same five phases. Connect the source and destination, scan to see what is actually there, rehearse the cutover so the weekend does not surprise you, migrate with live visibility for everyone watching, then monitor or roll back if needed. Here is the full walk-through.
Connect the source and destination
Pick the source platform (Google Drive, Dropbox, SharePoint, Box, SMB file share, and seven more) and the destination. MigrationFox supports OAuth, service-principal, API key, and Windows-agent authentication depending on what the platform offers.
Credentials are encrypted with AES-256-GCM at rest. Your file data streams directly from the source to the destination — MigrationFox never stores user content on its own servers.
What you’ll see
- Scope diff UI — for Graph destinations, the product tells you exactly which scopes are missing before you try to run. No guesswork.
- Credential test runs on every save. If a scope is wrong, you find out at configuration time, not at cutover time.
- Windows agent for on-prem sources (SMB file shares). No inbound firewall ports; the agent polls for jobs.
Discover what is actually in the source
Before a single byte moves, the discovery scan enumerates the source and builds a complete inventory — file count, size distribution, folder depth, permission mapping, and a junk-file auto-skip pass that trims operating-system garbage.
- Junk auto-skip filters
.DS_Store,Thumbs.db,~$lock,thumbdata3,.Trash-*,.Spotlight-V100, and a few dozen more. Typical drives shed 30–50% of their file count this way. Full reference. - Permission map lists every principal referenced in source ACLs, flags the ones that will not exist on the destination, and lets you upload a mapping CSV.
- Dependency graph (optional) scans Power Automate flows, Power Apps, and Teams tabs that reference the source site, so you know what depends on what you’re about to move. Read why.
Rehearse the cutover
Migration Rehearsal runs a full dry-run of every phase of the real migration against the real source, with destination writes short-circuited. It emits a verdict pill, a runtime prediction, a Graph API request-count forecast, and a 10-category issue report. Any problem the real cutover would hit, the rehearsal hits too.
- Three-state verdict:
ready/proceed with caveats/do not migrate. Designed for a change ticket, not a spreadsheet. - Ten issue categories:
schemaMismatch,permissionRisk,blockedFileType,oversizedFile,apiThrottleRisk,invalidCharacter,pathTooLong,unsupportedFeature,configurationGap,dataIntegrity. Reference. - Client sign-off record locks the rehearsal version against the real migration. If anything drifts, the real run flags it.
Run the migration
The real transfer. Many concurrent file streams with auto-split partitioning for large jobs. Large files use chunked uploads for reliability. Failed items are auto-retried with exponential backoff. Share a branded live-view link with the stakeholder so they watch the run without texting you.
- Client Live View — 30-day JWT-signed URL, no login, branded with the client’s logo. Revocable with one click.
- SPMT True History Mode — preserve original
Created/Modified/CreatedBy/ModifiedBymetadata on SharePoint destinations. - Resumable — if a job is interrupted, restart picks up where it left off. No re-transferring files that already completed.
Keep monitoring after the migration
A migration is not done when the bytes land — it is done when the new tenant has the governance posture you want. Continuous monitoring schedules any assessment to re-run monthly, tracks the composite score over time, diffs findings release to release, and emails you the moment a new Must-Do-Before-Copilot finding appears.
- Trend chart — composite score plotted over every run. Regression gets noticed.
- Finding diff — new, resolved, and persistent findings broken out separately.
- Alert emails — the moment a new Must-Do finding appears.
Roll back if something lands wrong
Every completed migration keeps a seven-day reversible window. Before you revert, a preview shows exactly what will be undone — file counts, permission removals, folder deletions — so you never guess what undo actually does.
- Seven-day window — long enough to notice a landing problem, short enough to keep the journal size bounded.
- Preview before commit — the preview is the audit trail. If it does not match the intent, cancel.
- Granular — roll back the whole run, a single library, or a specific permission change. Docs.
Ready to run the real thing?
Start with 2 GB free. No credit card required. Every connector, the discovery scan, the free governance snapshot, and the Windows agent are included.
Start Free → Browse use cases