How to Plan B-Roll (2026 Guide to Shot Lists That Actually Cut)

Key Takeaways

• B-roll does four jobs. Illustrate, cover, breathe, establish. Naming the job before you shoot is the whole method.
• Derive the list from the script, not from the location. Cut points come first; shots come from cut points.
• Cover is the mechanical reason b-roll exists. Without it, every edit in your main footage is a visible jump cut.
• Shoot three versions and hold longer than feels sensible. Ten seconds minimum, wide, medium, and detail.
• Quantity is predictable. Roughly one usable shot per twenty to thirty seconds of finished video, plus a margin.

The Four Jobs of B-Roll

A shot without a named job is a shot you will not be able to place. These are the four, and they have different requirements.

Illustrate

Showing what is being said. The claim is "the workshop still uses the original machines," so you show the machines. The claim is "the graph went up," so you show the graph.

What it requires: specificity. Illustrating b-roll must show the actual thing being discussed, which is why generic footage fails at this job more than any other. A stock shot of a generic factory does not illustrate a statement about this workshop.

Where it fails: vagueness. If the viewer cannot tell that this shot is the thing that was just mentioned, it is not illustrating, it is decorating.

Cover

Hiding an edit in the main footage. This is the mechanical reason b-roll exists. When you remove a stumble, a pause, or thirty seconds of tangent from a talking head, the two remaining pieces do not match and you get a visible jump. Cutting away to b-roll across the join makes the edit invisible.

What it requires: availability at the right moment, and enough length to cover the join comfortably. Covering shots do not need to be beautiful, they need to exist.

Where it fails: running out. The single most common b-roll shortage is not having anything to cut to at the exact moment you need to remove something.

Breathe

Giving the viewer a beat after something dense or emotional. A held wide shot with no dialogue after a difficult admission is doing real work, and cutting it feels like tightening while actually making the piece harder to watch.

What it requires: stillness and duration. Breathing shots are usually the longest ones in the edit and the ones inexperienced editors trim first.

Where it fails: being too busy. A shot with movement or detail demands attention rather than releasing it.

Establish

Orienting the viewer in place and time. Where are we, what kind of place is this, is it morning.

What it requires: width and legibility. Establishing shots are usually wide and usually early in a sequence.

Where it fails: arriving late. An establishing shot that appears after the viewer has already worked out where they are is redundant.

The reason to name the job is that it changes what you shoot. A covering shot can be almost anything as long as it exists. An illustrating shot has exactly one correct subject. Shooting without the distinction gets you many of the first kind and none of the second, which is why sequences fall apart at precisely the moments that matter.

Deriving a B-Roll List From the Script

The method, in order.

Read the script and mark every cut point. Anywhere you will remove something, compress a tangent, or join two takes. On a scripted talking head these are predictable. On an interview-led piece, mark them after the log rather than guessing.

Mark every claim that needs showing. Anywhere the script asserts something the viewer would benefit from seeing. These become illustrating shots and they are the ones with no substitutes.

Mark the pressure points. After the densest or heaviest passage, note where the viewer needs a beat. These become breathing shots.

Mark each new location or time shift. These become establishing shots.

Convert each mark into a shot list line. Subject, framing, movement or locked off, approximate duration, and the job it serves. A line reading "detail, hands on the lathe, locked off, 15s, illustrates the craftsmanship claim at 02:40" is actionable. A line reading "workshop b-roll" is not.

Group the list by location, not by script order. You will shoot everything in one place at once, and a list ordered by script order will send you back and forth.

Add a covering surplus. Beyond the specific shots, add a handful of general-purpose covering options per location, because cut points multiply during editing and you cannot predict all of them.

How Much B-Roll Do You Actually Need

The question everybody asks, with a usable answer.

Roughly one usable shot per twenty to thirty seconds of finished video. A ten-minute video needs somewhere around twenty to thirty usable b-roll shots. This is a starting estimate, not a rule, and it moves with format.

Shoot two to three times what the list says. Not more locations, more versions of the same subjects: different focal lengths and angles. This costs little on the day and saves the edit constantly.

Hold every shot for at least ten seconds. Longer than feels necessary while you are standing there, and never enough when you are editing. A shot that runs out before the sentence it is covering has ended is useless.

Formats vary predictably. A tutorial or explainer is heavy on illustrating shots because it makes claims constantly. A vlog is heavy on establishing and breathing. An interview-led documentary is heavy on covering, because you are removing a great deal of speech.

The honest test: if you have to use a shot twice in the same video, you were short. Occasional reuse is fine; noticeable reuse reads as a shortage and viewers register it even when they cannot name it.

Shooting B-Roll That Actually Cuts

Three versions of everything: wide, medium, and detail. This trio cuts together as a sequence on its own, which means any subject can become a five-second run rather than a single shot.

Hold still, or move deliberately. Slow drifting handheld with no intention is the least usable footage there is. Either lock it off or make the movement a decision with a beginning and an end.

Start and end the shot before and after the action. Roll for two seconds before anything happens and keep going for two seconds after. Editors need handles.

Shoot the same subject at two different moments if lighting allows. Cheap insurance against a continuity mismatch with the main footage.

Record ambient audio everywhere. Room tone and location atmosphere are the difference between b-roll that sits in the mix and b-roll that arrives as an audio hole. Thirty seconds of clean ambience per location, always.

Get the detail shots people forget: hands, textures, objects at rest, things being put down, a door opening. These cover almost anything and give an edit texture.

Avoid faces you have not cleared, particularly in public places, and get releases where you need them.

B-Roll for Talking Head and YouTube

The most common use case, with its own logic.

Cover is the dominant job. Most YouTube editing removes a large proportion of what was said, and every removal needs covering. Plan more covering shots than feels reasonable.

Illustrate anything you claim. Retention improves when the viewer sees the thing under discussion rather than a person describing it, and this is the mechanism, not a mystical engagement trick.

Screen recordings count as b-roll, and for software or tutorial content they are usually the primary b-roll source. Plan them as shots with subjects and durations, not as an afterthought.

Cut to b-roll before the viewer gets bored, not after. The standard practice of changing the image every several seconds is really a way of ensuring you never sit on a static frame long enough for attention to drift.

Do not cut away during the emotional or important line. The face is doing the work there, and cutting away throws it out. Cover the setup and the aftermath, not the moment.

Documentary B-Roll Is Evidence, Not Decoration

Worth separating, because the standard advice actively misleads here.

Illustrating shots carry the argument. When a contributor claims something, the shot that shows it is doing evidential work. This is the difference between a film that asserts and a film that demonstrates.

You often cannot restage it, so the shot list is a list of things to look for rather than things to set up. Documentary b-roll planning is closer to a hunting list than a shot list.

Shoot the ordinary things immediately. The routine details of a place are what you will need and are the easiest to assume you can get later. You frequently cannot; access changes.

Observational footage of people doing things beats abstract texture for nearly every documentary purpose, because it carries information as well as atmosphere.

Archive and stills are b-roll too, and their availability should be established during research rather than discovered in the edit.

Sourcing: Shoot, Stock, Archive, Generate

Shoot it yourself where the shot must be specific. Anything illustrating a particular claim about a particular subject has to be real.

Stock is acceptable for establishing and breathing, where the requirement is a category of image rather than a specific one. Check the licence covers your use, including monetised video.

Archive footage needs clearance and often payment, and the lead time is real. Establish availability during research.

Generated footage is now viable for certain abstract and illustrative shots. Two honest cautions: platform and client disclosure expectations are changing quickly, so check what applies to you, and generated footage cannot serve the evidential job at all, because the entire value of an illustrating shot in factual work is that it is real.

Editing With B-Roll

Keep the original audio underneath. Cutting to b-roll while the interview or voiceover continues is the standard construction. The b-roll's own audio usually sits low in the mix as texture rather than replacing anything.

Cut on the sentence, not mid-phrase, unless you are deliberately covering a join, in which case cut where the edit is.

Let establishing shots run. They are doing orientation work and clipping them to a beat defeats the purpose.

If a shot has no job, remove it. A shot included because it is beautiful and not because it is needed is the most common cause of a video that feels padded.

Quick Picks: Tools for Planning B-Roll

Best for a formal shot list: StudioBinder. Purpose-built shot listing with framing, movement, and equipment fields, tied to a schedule. The right choice on a crewed production. Free tier, paid from around $29/month (verify current pricing).

Best for deriving the list from the script: Storyflow. The canvas where the script sits beside the b-roll list, so each shot is attached to the cut point or claim that generated it and the AI can read both together. Free plan is $0; Plus is $7.99/month billed annually. The honest limit: it is not a shot list format anyone on a crewed set expects, and it does not schedule.

Best free option: a spreadsheet. Columns for subject, framing, duration, job, and location, sorted by location for the shoot day. What most creators actually use, and sufficient.

Best for solo creators already working in a notes app: Notion or Apple Notes. Lower friction than any dedicated tool, and friction is what actually stops people from planning b-roll at all.

Pricing checked July 2026. Competitor pricing changes often and varies by plan and region, so verify on each vendor's page before quoting it.

Common Mistakes

Shooting without naming the job. Produces footage that fits anywhere and is needed nowhere.

Too few covering shots. The most common actual shortage, discovered at the worst time.

Shots that are too short. Ten seconds minimum. Editors need handles.

Only one version of each subject. Wide, medium, and detail cost thirty extra seconds and save the edit repeatedly.

No ambient audio. B-roll arrives as an audio hole and the mix suffers.

Cutting away during the important line. The face was doing the work.

Reusing the same shot several times. Reads as a shortage even when the viewer cannot name it.

Using stock for illustrating shots in factual work. It does not illustrate, it decorates, and in documentary it quietly undermines the claim.

The Bottom Line

B-roll planning fails for one reason, and it is not laziness. People shoot without deciding what each shot is for, because deciding is harder than filming, and attractive footage feels like progress.

Name the job first. Illustrate, cover, breathe, establish. Derive the list from the script's cut points and claims rather than from what looks good at the location. Shoot three versions, hold them longer than feels sensible, and record the room.

Do that and the edit becomes assembly rather than rescue. Skip it and you will have a drive full of good-looking shots and nothing to cut to at 04:12, which is the only moment that mattered. Generic B-roll is footage that fits anywhere and is needed nowhere.

Author

Justkay is a documentary filmmaker and the founder of Storyflow. He has shot b-roll on documentary projects where a sequence depended on having the one specific insert that proved what a contributor had just said.