How to Make a Documentary: Finding, Shooting, and Structuring

The short version: find a person who wants something badly, shoot more than you think you need, and accept that the real movie gets built in the edit, not on the day. If you are learning how to make a documentary, that sentence is about 80 percent of the job.

The other 20 percent is not quitting when the story refuses to show up on schedule.

The day the story refused to cooperate

Early in my career I showed up to shoot a profile piece with a shot list, interview questions, and a clean three-act structure in my head. By lunch, the subject had told me his real story was something else entirely. The thing I had planned to film was the press release version of his life. The actual story was quieter, messier, and ten times more interesting, and it only came out because the camera was still rolling when the "real" shoot ended.

Fifteen years of cutting films taught me that this is not an accident. This is how documentaries work. Fiction starts with a script and bends reality to fit it. Documentary starts with reality and earns the script afterward. You do not find the story by planning harder. You find it by showing up, paying attention, and staying long enough for the rehearsed answers to run out.

A film critic friend of mine once broke down Joker, a 2019 Hollywood movie, and pointed out it runs on what he called a weak or even non-existent plot. The movie works anyway, he argued, because it is anchored to one character the audience understands completely. That is the documentary filmmaker's entire situation. You rarely have a plot when you start shooting. You have a person. If the person is compelling enough, with a clear want and real obstacles, the plot will arrive. If the person is not, no amount of pretty footage will save you.

You make the documentary twice

Here is the part nobody tells beginners: a documentary is made twice, once in the field and once in the edit.

The first version is the one you shoot. You chase the person, the place, the event. You collect. You are a detective with a camera, and your only job is to not miss the moment when the mask slips, because that moment is the movie.

The second version is the one you build in the edit room, and this is where most documentaries are actually won or lost. In the edit you ask the brutal questions: what does this person want? What is stopping them? Why should a stranger care? One breakdown I studied of John Carter put it perfectly: if you do not first clearly establish the importance of what is going on, the audience will not care about what is going on. That applies to a documentary audience exactly the same. Nobody owes your film their attention. You have to make the stakes visible, and you have to do it early.

The edit is also where you discover what your film is actually about, which is often not what you thought when you started. This is normal. Fight it and you will force a story that does not exist. Follow it and the film gets honest.

How to make a documentary shoot work: three field rules

Shoot lean, but shoot a lot. Lean does not mean lazy. It means a small crew, minimal gear you can carry all day, and the discipline to keep rolling through the boring parts. The boring parts are where people stop performing. Audio matters more than the camera. Audiences will forgive a shaky frame; they will not forgive unusable sound. I have watched good stories die because nobody monitored the levels.

Chase the want, not the topic. "A film about a fishing village" is a topic, and topics are boring. "A film about a fisherman trying to keep his boat before the bank takes it" is a story, because someone wants something and something is in the way. One analysis of Dead Man's Chest I love praised it for being a goal and motivation-driven narrative. Fiction or documentary, that engine is the same. Before every shoot day, write down one sentence: what does my subject want today? If you cannot answer it, you are filming wallpaper.

Show, never tell, especially in interviews. The same Joker breakdown made a point I think about constantly: showing is always more powerful than telling. In documentary terms, this means a subject crying on camera is worth ten minutes of someone explaining that times were hard. It means the scene of the eviction beats the interview about the eviction. Plan your shoots around events and actions, not just talking heads. Interviews are the glue. Observational footage is the film.

Key points

• A documentary is made twice: in the field and in the edit. Budget your time for both.
• Start with a person who wants something, not a topic. Topics do not have stakes; people do.
• Establish why the audience should care early. Nobody owes your film their attention.
• Shoot lean: small crew, reliable audio, and keep rolling past the rehearsed answers.
• The edit reveals the real film. Be willing to abandon the story you planned for the one you found.
• Build shoots around actions and events, not interviews. Showing beats telling.

Try this

• Write one sentence before your next shoot: what does your subject want this week? Tape it to your camera bag.
• Watch the opening ten minutes of any documentary you admire and time how long it takes before you understand what the main person wants. Notice how the good ones answer it fast.
• Take one interview you have already shot and cut a 60-second version using only observational footage, no talking heads. See what survives.