Before you book a reshoot, describe the shot in words someone else could recognize. Then search the places footage ends up, cheapest first: finished videos, then the editor's project files, then the raw footage. Most checks stall at the raw footage, because it is rarely logged shot by shot. If that's where you are, a search layer that indexes what's in the video lets you ask for the shot by description, so you don't have to scrub through it. Decide one thing before you start. The check is over when you can say what you searched and how, because "we didn't find it" and "we don't have it" are different answers.
Write the shot down first
"Customer testimonial" is a category. "A customer at a whiteboard explaining the problem, medium shot, our old office in the background" is a shot, and that's the version you can check footage against. Also write down what would rule a clip out even if you found it, like an old logo or a product screen that has since changed. People matter too: someone who has left the company may have signed a release that doesn't cover a new use. Plenty of reshoots happen because the footage exists but fails one of these tests. You want to know that before the search, not after it.
Look in order of cost
Start with published and exported videos. If the shot ever made it into a finished piece, like a launch video or an event recap on the company channel, you can trace it back to the edit that used it.
Next, ask whoever cut those videos. An in-house editor or an outside agency usually has the project files, and project files point to the source clips. This is often faster than any search, because one person remembers the shoot.
Raw folders come last, and they're the hard part. Camera filenames say nothing about content, and folders are sorted by date or shoot, not by what's in them. When a team builds a shared library for other staff to pull from, this layer decides whether anyone actually uses the library.
Ways to search raw footage
Scrubbing by hand costs nothing but time. It works when there's one shoot to check and stops working when the answer could be "somewhere in the last two years of events."
Tags in a DAM work if someone logged the footage when it came in. The catch is that tags record what the logger thought mattered back then, which often isn't the shot you need today. How far this gets you depends on how the team structured its asset library in the first place.
Transcript search works when the shot is defined by something said on camera. It can't find a shot that is only visual, like office b-roll or a product sitting on a desk. Searching by spoken words covers that route in detail.
The fourth option is a content search layer. You upload the footage, it gets indexed once, and then you describe the shot. You pay for the upload step and a usage plan, and you still judge every result yourself. With Vivu, footage from a campaign or event goes into a project and is indexed once in the cloud. After that, "do we already have a close shot of the product on a desk" is a question you type into that project, and you get back time ranges to open and watch before you decide. Each search covers one project, so if past shoots are spread across several projects, pick the one a pre-shoot check should run against.
When a search comes back empty
An empty result is ambiguous whatever tool you use. It can mean the shot isn't in the footage, or it can mean the description was stricter than what was filmed. An event recorded mostly from a wide stage angle may have no close shot of a single speaker, even though the wide version is there and might crop well enough. Before you call it a reshoot, loosen the description. Drop the framing, search for the scene, and check the angles by eye. If the looser search also comes back empty, you have a much better reason to book the crew.
When you don't need any of this
If one person shot everything and still remembers it, ask them. If the shot is cheap to recreate, like a phone clip of the product or a screen recording, reshooting can cost less than an hour of looking. And if the brand or product has changed since the footage was filmed, old clips will fail your checklist anyway. The search would only tell you what you can't use.
Which side you're on
If "do we have this?" comes up a couple of times a year, following the order above by hand is enough. If it comes up at the start of most shoot plans, and the answer depends on someone remembering a folder, the problem is how findable your raw footage is. That's the point where a searchable library pays back its setup.
FAQ
How long should we search for old footage before deciding to reshoot?
Set the limit before you start, based on what the reshoot would cost. A simple phone or screen shot rarely justifies a long search. A shot that needs talent or a customer's time usually does. Write down what you checked so the next person doesn't repeat the same search.
Can we reuse footage that shows a former employee or an old logo?
Sometimes. For people, check the release they signed and whether it covers the new use. That's a question for whoever holds the paperwork, often legal or the producer. An old logo or an outdated product screen usually rules a clip out for external use, but the clip may still work internally.
Our agency shot the footage. Do we actually have it?
Only if you received the raw files, or the contract says the agency keeps them for you. Ask for the project files and source footage, and check the contract for who owns the raw material. Many teams find out they only have the final exports.
Is a DAM enough to tell us whether we already have a shot?
It is if the footage was tagged shot by shot when it came in. Most libraries are tagged by file or by shoot. That tells you a shoot happened, but not whether a specific angle or moment is in it, so you still have to open the files or search the content.