Every wedding videographer eventually faces the same quiet question: the film is delivered, the final payment has cleared, and there are 400 GB of footage still sitting on a drive. How long do you keep it? Where does it live? And who decides when it is finally safe to delete? Most studios never decide. The footage simply accumulates until a drive fails or a client emails three years later asking for "just one more clip from the first dance."
A retention policy is the answer to all of those questions, written down once and applied to every wedding. It protects your clients' memories, protects you from liability, and keeps storage costs predictable. It also matters for anyone who hands footage to an outsourced editor, because every editor you work with holds a copy of client material that needs a defined end date. This guide covers what to keep, for how long, in which tier, and how to word it in your contract.
Why You Need a Written Retention Policy
A written retention policy turns a vague habit into a business rule, and that single change removes most of the risk around old footage. Without one, every request for old material becomes a custom decision, and every decision is made under pressure.
The three problems an unwritten policy creates are predictable. First, storage costs creep: drives pile up in a cupboard, each one a small unmanaged liability. Second, expectations drift: clients assume you keep their footage forever, while you assume it is gone after delivery. Third, privacy exposure grows, because footage of guests, children and private ceremonies exists on devices you no longer track.
A policy fixes all three by answering four questions in advance: what categories of material you hold, how long each category is kept, where each category lives, and what happens at the end of the period. Clients who read this in your contract rarely object. Most simply want to know you have a plan.
| Situation | Without a policy | With a policy |
|---|---|---|
| Client asks for a clip two years later | You scramble through drives, unsure it still exists | You check the archive index and know exactly where it is |
| Drive fails | Unknown what was only on that drive | Every project has a defined second copy |
| Editor leaves your team | Their copies of client footage linger indefinitely | Deletion dates and confirmations are already agreed |
| Client disputes what was promised | You rely on memory | The contract states the retention window |
What Counts as Footage Worth Keeping
Not all material from a wedding has the same value, and treating it equally is what makes archives expensive. Sort every project into four layers, because each layer deserves a different retention decision.
The first layer is the delivered films: the highlight, the full-length edit, any teasers and the social cuts. These are small, irreplaceable in their finished form, and what the client actually paid for. They are kept the longest.
The second layer is the project files: the Premiere, Resolve or Final Cut project, the sequences, grading data and relinking information. A project without its media is of limited use, but a project file plus the source media lets you re-export in a new format years later. If you work with an outsourced editor, agree on who holds the project file and for how long.
The third layer is the original camera files: the untouched card dumps. These are the bulk of your storage and the hardest to replace. They are the only way to re-edit the wedding from scratch, so they are the layer your policy needs to treat most carefully.
The fourth layer is derivatives and scratch material: proxies, render caches, conformed audio and exports of work-in-progress cuts. These can be regenerated and should be deleted early, usually when the project is delivered. Our guide to the proxy workflow explains why proxies belong in this disposable layer.
| Layer | Typical size | Replaceable? | Suggested treatment |
|---|---|---|---|
| Delivered films | Small (a few GB) | Only by re-exporting | Keep long-term, multiple copies |
| Project files | Very small | No | Keep alongside the originals |
| Original camera files | Largest (hundreds of GB) | No | Keep for a defined window, then decide |
| Proxies, caches, scratch exports | Large | Yes | Delete after delivery |
Retention Windows: How Long Is Long Enough
There is no legal rule that sets a retention period for wedding footage in most places, so the right window is a business decision balanced against cost and client expectation. The most common pattern among working studios is a short "hot" window after delivery, a longer "cold" window for originals, and indefinite storage for the finished film.
The hot window covers the period when revisions, re-exports and reprint requests are most likely. Many studios keep full working copies for roughly the first few months after delivery, which covers the revision rounds described in our guide to structuring revision rounds. After that, requests drop sharply.
The cold window covers the originals. A common choice is one to three years after delivery, which comfortably covers anniversary requests, album and print projects, and the occasional dispute. Some studios offer longer retention as a paid add-on, which turns an unavoidable cost into a small revenue line.
The finished film is different. It is tiny compared with the originals, and it is what clients return for. Keeping delivered films indefinitely costs almost nothing and removes the most common source of "can you resend it" emails.
| Stage | Example window | What lives here | Purpose |
|---|---|---|---|
| Hot | Delivery to a few months | Originals, project, proxies on fast storage | Revisions, re-exports |
| Cold | Up to a defined number of years | Originals and project on archive tier | Late requests, disputes |
| Permanent | Indefinite | Delivered films, a client copy link | Resends, anniversaries |
| Deleted | After window ends | Nothing | Privacy and cost control |
These windows are examples, not rules. Pick numbers you can actually honor, write them down, and then follow them. A modest promise you keep is worth more than a generous one you quietly abandon.
Where Each Tier Should Live
Match the storage to the tier: fast drives for the hot window, a cheaper and slower option for the cold window, and at least two independent copies of anything you cannot recreate. The principle behind this is the same one covered in our guide to the 3-2-1 backup rule: three copies, on two kinds of media, with one copy off-site.
The hot tier is a fast external SSD or RAID that you edit from. It is not an archive, and it should never be the only place the originals exist.
The cold tier is where cost matters most. Two practical options dominate for small studios: offline hard drives that are labeled, indexed and stored somewhere safe, and cloud object storage. Cloud storage removes the problem of drives failing in a drawer, but you pay monthly forever. Offline drives cost more up front but nothing monthly, and they require you to check them periodically.
For scale, Backblaze reported an annualized hard drive failure rate of 1.36% across its fleet in 2025. That is a low rate for a large fleet, but it is a reminder that a drive sitting alone in a cupboard is not protected by anything except luck. A single copy of irreplaceable footage on a single drive is a gamble, even a good gamble.
| Option | Upfront cost | Ongoing cost | Main risk | Best for |
|---|---|---|---|---|
| Fast external SSD or RAID | Moderate | None | Failure, theft, being reused | Hot tier only |
| Labeled offline hard drives | Moderate per drive | None | Drive failure, forgotten or lost | Cold tier with a second copy |
| Cloud object storage (e.g. Backblaze B2) | None | Monthly per TB | Recurring cost, account access | Off-site copy |
| Cold cloud archive (e.g. AWS S3 Glacier Deep Archive) | None | Very low per TB, plus retrieval fees | Slow restores, minimum storage periods | Rarely accessed originals |
Whichever you pick, keep the delivered film and the project file in a more accessible place than the bulky originals. They are small enough to keep in ordinary cloud storage and they are what you will actually need most often.
What Archiving Actually Costs
For a typical studio, archiving costs far less than most videographers assume, and the real cost is usually the time spent organizing it rather than the storage itself. It is worth doing the arithmetic once so the policy is based on numbers instead of fear.
Published pay-as-you-go cloud prices give a useful sense of scale. As of autumn 2026, Backblaze B2 lists storage at roughly $6.95 per TB per month, with downloads free up to three times your average monthly storage. AWS S3 Glacier Deep Archive lists storage at roughly $0.99 per TB per month in the US East region, with a 180-day minimum storage duration and retrieval times measured in many hours. Prices change, and fees such as retrieval and requests are billed separately, so always confirm on the provider's own pricing page before you commit.
Take a studio that shoots 30 weddings a year at about 400 GB each. That is roughly 12 TB of new originals annually. On a B2-style plan, holding one year of originals costs on the order of $80 a month. On a deep-archive tier, the same data costs a small fraction of that, but retrieval is slow and billed. Neither number is large compared with the price of a single wedding package.
| Scenario (12 TB of originals) | Approximate storage cost | Retrieval experience |
|---|---|---|
| Standard cloud object storage at about $6.95/TB/month | About $83 per month | Immediate, with a generous free download allowance |
| Deep archive at about $0.99/TB/month | About $12 per month | Many hours, with retrieval fees |
| Offline drives, two copies | One-time drive purchase | Immediate, but you must keep them safe and tested |
The right answer is often a blend: keep the first year of originals somewhere you can reach quickly, and move older originals to the cheapest tier once requests become rare. The numbers above are illustrations of scale, so check live pricing before budgeting.
Indexing: The Part Everyone Skips
An archive you cannot search is only a pile of hidden files, and indexing is the step that makes retention worth the effort. The test is simple: could you find a specific wedding's originals in under five minutes, without opening a single drive to look?
Keep one master spreadsheet. For each wedding, record the couple's names, the wedding date, the project name, where the originals live, where the second copy lives, where the delivered film lives, the date originals are due for review, and the date they are due for deletion. This one file turns a retention policy from an intention into a system.
Consistent folder names are what make the index reliable. Name every project the same way, for example date, then surname, then a short project label, and use the same structure on every drive. Label every physical drive on the outside with a unique ID that matches the spreadsheet.
| Index column | Example entry | Why it matters |
|---|---|---|
| Project ID | 2026-06-14_Smith | Matches folder and drive labels |
| Originals location | Drive A07, cloud bucket | Tells you where to look first |
| Second copy | Drive B07 | Proves the backup rule is met |
| Delivered film link | Cloud folder URL | Fast answer to resend requests |
| Review date | 2028-06-14 | Triggers the keep or delete decision |
| Deletion confirmed | Yes or no, with date | Closes the loop for privacy |
Privacy, Contracts and Outsourced Editors
Retention is partly a privacy question, because wedding footage contains guests, children and private moments, and holding it longer than you need increases your exposure. The simplest protection is to say, in your contract, what you keep and when you delete it.
Your contract should cover four points: the retention period for originals, the retention period for the finished film, how a client can request a copy or early deletion, and that retention periods may end earlier if storage is lost through no fault of yours. Have a lawyer in your own jurisdiction review the wording, since data protection rules differ between countries and regions. This is general information, not legal advice.
If you outsource editing, the same discipline has to extend to your editor. Every editor receives full copies of client footage, and each of those copies needs an agreed end date. Our guide to vetting an outsourced editor covers the questions to ask, and deletion confirmation belongs on that list. Ask for written confirmation once the project is delivered and approved, and record it in your index.
| Who holds a copy | What they hold | Agreed end point | Confirmation |
|---|---|---|---|
| You | Originals, project, films | Per your policy window | Index entry |
| Outsourced editor | Originals or proxies, project files | After final approval, within an agreed number of days | Written confirmation |
| Client | Delivered film | Theirs to keep | Delivery record |
| Cloud provider | Whatever you uploaded | Your deletion date | Bucket or folder cleared |
Running the Archive Process and the Final Decision
The archive process should run the same way after every wedding, because consistency is what prevents the gaps that cause loss. Build it as a short checklist that takes less than an hour per project once you are used to it.
Start the moment the client approves the final film. First, export and store the final deliverables in every agreed format. Second, save the project file and confirm it opens and relinks to the originals. Third, copy the originals to the archive tier and verify the copy with a checksum or a file-count and size comparison, not just by looking at the folder. Fourth, add the second copy. Fifth, delete the proxies, caches and scratch exports. Sixth, update the index with locations and the review date. Seventh, request deletion confirmation from any editor.
Verification is the step most often skipped, and it is the one that saves you. A copy that was never checked is only a hope. Run a verify pass before you erase anything from the hot tier, and never delete the hot copy until the cold copy has been confirmed.
| Step | Action | Done when |
|---|---|---|
| 1 | Store final deliverables | All agreed formats saved and playable |
| 2 | Save and test the project file | Opens and relinks cleanly |
| 3 | Copy originals to cold tier and verify | Checksum or size and count match |
| 4 | Create the second copy | Independent location confirmed |
| 5 | Delete proxies and scratch | Hot tier cleared of derivatives |
| 6 | Update the index | Locations and dates recorded |
| 7 | Request editor deletion | Written confirmation filed |
A worked example shows how the pieces fit. Imagine a two-person studio that delivers a wedding in early June. On delivery day, the highlight film, the full film and the social cuts go into a permanent client folder, and the client receives a link. The proxies and render caches are deleted within the week. The originals stay on fast storage through the hot window, then move to the cold tier with a verified second copy, and the index records a review date two years out. When that date arrives, the studio checks for pending requests, then deletes, extends or hands over according to the contract. Nothing about this requires special software; it requires a calendar, a spreadsheet and the habit of finishing the checklist before starting the next edit.
It also helps to decide who owns the process. In a solo studio that is you, but in a team it should be one named person, because shared responsibility for archives usually means nobody does it. Put the review dates in a recurring calendar reminder, so a deletion decision arrives as a prompt rather than something you must remember. A short quarterly session, where you work through every project whose review date has passed, keeps the whole system from silently falling behind.
Reviewing and Deleting: The Final Decision
When an originals window ends, you have three honest choices: delete, extend, or hand the footage to the client. Making that choice deliberately, on a calendar date, is what separates a policy from a pile.
Deleting is appropriate when the window has passed, the film is safely delivered, and the client has not asked for an extension. It reduces cost and privacy exposure. Do it in a way you can document, and record the date.
Extending is appropriate when the client has asked, when an album or print project is still in progress, or when the footage has ongoing portfolio or business value and the couple has given permission to use it. A paid extension is a fair way to cover the continuing cost.
Handing over is an option some studios offer for a fee: the client receives the originals and the studio deletes its copies. This only works if your contract allows it and the client understands they take on the responsibility of storing the footage safely.
| Choice | When it fits | Watch out for |
|---|---|---|
| Delete | Window ended, no request pending | Make sure the second copy and finished film are safe first |
| Extend | Client request or ongoing project | Agree the fee and the new end date in writing |
| Hand over | Client wants full ownership of originals | Contract language, verified copy, then delete your own |
Be equally clear about what the client can expect to receive. Many couples assume that paying for a film includes the raw footage, so state in the contract whether originals are delivered, stored for a window only, or available for a fee. Ambiguity here is the usual root of disputes, and a single sentence prevents them. The same applies to the finished film: say how long the download link stays active, and what you do if the client loses their copy. A permanent, low-cost hosted copy of the finished film is a small investment that answers most late requests before they become emails.
Common Mistakes and How to Avoid Them
Most archive failures come from a handful of avoidable habits, and each one has a simple fix. Knowing them in advance is the cheapest protection you can buy.
The first mistake is relying on one drive. A single copy is not an archive, no matter how new the drive is. Always keep an independent second copy of anything you cannot recreate.
The second is never checking the copies. Drives sit in drawers, accounts lapse and files go corrupt silently. Schedule a check, even once a year, where you open a sample of files from each cold drive and confirm they play.
The third is keeping everything forever by default. It feels safe, but it grows cost, clutter and privacy exposure. A defined deletion date is a feature.
The fourth is forgetting the editor. Outsourced editors and second shooters hold real copies of client footage. Include them in your policy and get the deletion confirmation in writing.
The fifth is deleting before verifying. Never clear the hot tier until the cold copy has been confirmed. The order of operations is the whole point.
| Mistake | Consequence | Fix |
|---|---|---|
| Single copy of originals | One failure loses the wedding | Independent second copy, ideally off-site |
| Never testing archives | Silent corruption found too late | Annual sample check |
| Keep everything forever | Rising cost and privacy exposure | Written deletion dates |
| Ignoring editor copies | Unmanaged client data elsewhere | Deletion confirmation in writing |
| Delete before verifying | Irrecoverable loss | Verify cold copy first, always |
One more habit is worth adopting: rehearse a restore before you need one. Once a year, pick a random past wedding and pretend a client has asked for a specific clip. Locate the originals using only your index, retrieve them from the cold tier, open the project file, and export a short test clip. Time the whole exercise and note anything that surprised you, such as a missing relink, a drive that would not mount, or a cloud login nobody remembered. A rehearsal costs an afternoon and exposes exactly the weaknesses that would otherwise appear during a real request, when the client is waiting and your reputation is on the line. Studios that run this test once usually fix their index and labeling within the same week.
Finally, keep the policy itself short enough to be used. A one-page document listing the four layers, the retention window for each, the storage location and the deletion rule is far more valuable than a long manual nobody opens. Share the relevant lines with any editor or second shooter you work with, and review the whole page once a year as your storage costs, camera formats and client expectations change.
Frequently Asked Questions
How long should a wedding videographer keep raw footage?
There is no universal rule, but one to three years after delivery is a common and defensible window for originals, with the finished film kept indefinitely. Choose a period you can honor, state it in your contract, and apply it to every client.
Can I delete the footage after the client approves the film?
You can, if your contract says so and the client has agreed. Most studios keep the originals for a defined window first, because late requests and revision questions do happen after approval. Delete the proxies and scratch files early, but be more careful with the originals.
Is cloud storage better than hard drives for archiving wedding video?
Neither is strictly better. Cloud storage removes the risk of a forgotten drive failing but costs money every month, while drives cost nothing ongoing but need labeling, safe storage and periodic testing. Many studios use both, which also satisfies the off-site part of the 3-2-1 rule.
What should my contract say about footage retention?
It should state how long you keep originals, how long you keep the finished film, how clients can request copies or early deletion, and that you are not liable for loss beyond the stated window. Have a local lawyer review the wording, because data protection rules vary by country.
What happens to footage my outsourced editor holds?
It should be deleted on an agreed date after final approval, and you should ask for written confirmation. Treat the editor's copies as part of your own archive policy, since they hold real client material.
Should I charge clients to keep their footage longer?
Many studios do, and it is a fair approach because storage and administration carry a real continuing cost. A modest annual retention fee, stated in the contract, turns an open-ended obligation into a clear option the client can choose.