Preparing a Media Archive Around the Questions You Will Ask Later
Most people build a media archive by saving files as they go and planning the structure later. The problem appears months afterward, when you need one specific clip or a set of images for a particular purpose and discover that the filenames, folders, and dates tell you very little. A better approach begins with a different question: what will you need to find, and how fast do you need to find it? Designing the archive around those future questions is usually more effective than designing it around the files you have today. This article walks through a method for doing that, along with a checklist you can adapt to your own collection. Nothing here is a promise about any platform's storage or recovery behavior; the goal is to build habits and structures that work regardless of where your files live.
Start with a list of retrieval questions
Before you organize anything, write down ten to fifteen questions you expect to ask of your media. Examples include: Which clips show a specific location? Which photos were taken in a particular lighting setup? Which files were already approved for a past post? Which versions are the final edits and which are drafts? These questions reveal the attributes your archive must capture. If you often ask about locations, location needs to be a recorded field. If you often ask about approval status, that status needs a place in your system.
Write the questions in plain language and rank them by how often you expect to need them. The top three or four should shape your folder structure and naming convention. Lower-frequency questions can be answered through search or notes rather than through dedicated structure. This keeps the system lean and prevents you from building categories you will never use.
Inventory before you restructure
A common mistake is to begin moving files before knowing what exists. Instead, create a simple inventory: a spreadsheet or plain text file listing each folder or batch, approximate file count, file types, date range, and the production stage it represents (raw capture, selects, edited, final, delivered). This inventory does not need to be perfect. Its purpose is to show duplication, gaps, and mixed-stage folders that would confuse any future search.
Once the inventory exists, note any batches that contain more than one stage. A folder labeled 'shoot March' that holds raw files, edits, and exports is a folder that will produce confusing retrieval results. Splitting it by stage, even as a temporary measure, improves clarity immediately.
Choose a retrieval model: folders, tags, or both
Folders work well when your archive has a clear hierarchy, such as project, then stage, then date. They are easy to browse and easy to back up. Their weakness is that a single clip often belongs to several categories at once, and forcing it into one folder hides the others. Tags solve this by allowing multiple attributes per file, but they depend on consistent entry and can become unmanageable if the vocabulary drifts over time.
A practical compromise is to use folders for stage and project, which change rarely, and tags or metadata fields for subject, location, mood, or usage, which change often. Keep the tag vocabulary short, write it down, and review it every few months. If you find yourself inventing a new tag for every shoot, the vocabulary is too granular and should be merged into broader terms.
Name files so they sort and read well
Filenames are the cheapest form of metadata you control. A consistent pattern such as date in ISO format, a short project code, a subject keyword, a sequence number, and a version suffix gives you sortable, readable names. Avoid spaces and special characters where your tools may struggle, and keep the pattern identical across the whole archive. Renaming a large collection later is tedious, so it is worth getting the pattern right at the start.
Version suffixes deserve particular care. Use a simple, monotonic scheme such as v01, v02, and reserve a final marker for the file you actually delivered. Without this, it becomes difficult to tell whether 'final_new_FINAL' is the approved version or just the latest attempt.
A checklist for auditing your archive
Use this checklist once a quarter, or before any large search or migration. First, confirm that every top-ranked retrieval question can be answered within two steps using folders, filenames, or tags. Second, check that each project folder contains only one production stage. Third, verify that version suffixes are consistent and that delivered files are marked. Fourth, review your tag vocabulary and merge or retire terms that are rarely used. Fifth, confirm that your inventory spreadsheet matches what is actually stored, and update it if not. Sixth, note any files whose origin or permission status you cannot confirm, and set them aside for review rather than leaving them mixed in.
Finally, write a one-page description of your archive's structure and keep it alongside the files. A future you, or a collaborator, should be able to read it and know where to look. Archives fail less often from missing files than from missing explanations.