Keeping every frame of raw footage from every project indefinitely isn't actually practical once you've been working for enough years, the storage cost and management overhead genuinely add up. I had to build an honest, deliberate system for what survives long-term archiving versus what gets deleted after a project wraps.
The tiered system that's actually worked
Final delivered files, graded and exported, get kept indefinitely, they're small relative to raw footage and the value of having them accessible for years is obvious. Selected, flagged raw footage, the frames actually used or genuinely strong alternates, gets archived long-term too. The vast bulk of unused, unflagged raw footage gets deleted after a defined window, typically a year past project delivery, once the realistic chance of needing it has passed.
Why "keep everything forever" isn't actually a strategy
Unlimited raw footage retention sounds responsible until you're paying for and managing petabytes of footage that will, statistically, almost never get revisited, and the false sense of security around "I still have the raw files somewhere" often doesn't survive contact with actually trying to find something specific in a genuinely disorganized multi-year archive anyway. A smaller, well-organized archive of what actually matters beats a massive, chaotic one of everything.
- Final delivered files: kept indefinitely, small and clearly valuable
- Flagged, used raw footage: archived long-term
- Unflagged, unused raw footage: deleted after a defined retention window
The client conversation this requires
I now state my retention policy clearly in project agreements so clients understand raw footage isn't kept forever by default, which has prevented a couple of awkward conversations where a client assumed unlimited access to raw material years after a project wrapped. Being upfront about the actual policy, rather than an implied but undefined promise of infinite storage, has made this a non-issue rather than an occasional point of friction.