B-Roll for Talking Head Videos That Keeps Viewers Watching

A talking head video can lose half its viewers in the first minute, and the face on screen is rarely the reason. The usual problem is that nothing changes visually for ninety seconds, so the eye wanders off before the argument lands.

Good b-roll for talking head videos fixes that without turning the piece into a slideshow. I cut interview and explainer content most weeks, and here is how I plan b-roll, shoot it, place it and trim it.

What b-roll is actually for in a talking head edit

B-roll has three jobs, and every clip you add should do at least one of them:

Show what the speaker is describing. If someone says "the strap frays after six months", show the frayed strap.
Cover an edit. When you remove a stumble or merge two takes, a cutaway hides the jump.
Reset attention. A change of image at the start of a new point tells the viewer a new idea has begun.

Clips that do none of these are decoration. Generic shots of a keyboard or a coffee cup feel like filler because they are filler.

Planning b-roll for talking head videos before the shoot

Most weak b-roll comes from shooting it last, when everyone is tired and the light has gone. I build a shot list straight from the script or interview outline instead.

Print the script or transcript and underline every noun you could physically show: a product, a place, a document, a hand doing a task.
Mark every spot where you expect to cut, such as a long tangent you plan to trim.
For each underline, write one wide shot and one close detail. A wide of the workshop and a close of the drill bit gives you two options for the same line.
Aim for roughly one usable cutaway per 20 to 30 seconds of final runtime. A 6 minute video needs around 12 to 18 shots, so shoot at least 25 to give yourself choices.

Shoot each clip for at least 8 to 10 seconds with no camera movement at the start and end. Those still handles give you room to trim and to add a dissolve if you need one.

Before you start cutting, put the b-roll in its own bins named by topic, not by card or date. The habits in organising a video edit before the first cut save more time here than anywhere else, because you will be hunting for specific shots dozens of times.

Where to place cutaways on the timeline

Put the talking head on video track 1 and b-roll on track 2 above it. That way you can slide, trim and delete cutaways without ever touching the interview edit underneath, and the audio from the speaker stays continuous.

Start the cutaway slightly after the cue word

If the speaker says "the old hinge", bring the hinge shot in 4 to 8 frames after the word begins, not before. Cutting in early makes the image feel like it is predicting the sentence. Cutting slightly late feels like the picture is answering it.

Return to the face for emphasis

Come back to the speaker for any line that carries opinion, emotion or a key number. Viewers trust a face more than an image when the claim matters. A simple rule: show the thing during description, show the person during judgement.

Cover cuts with a full sentence, not a fragment

When b-roll hides a jump, let it run across the edit point and finish on a natural pause. A cutaway that starts and ends inside the same sentence draws attention to itself. If you are trying to decide which cuts need covering at all, the guidance on jump cuts in talking head videos covers when an honest jump cut works better.

How long each clip should stay on screen

Duration is where most edits go wrong. These are the ranges I start from, then adjust by feel:

Shot type	Typical length	Why
Close detail of an object	2 to 3 seconds	Viewers read it almost instantly
Hands performing a task	3 to 6 seconds	The action needs to complete
Wide establishing shot	3 to 4 seconds	Long enough to orient, short enough not to drift
Screen recording	5 to 12 seconds	The viewer may need to read text or follow a click
Sequence of several shots	8 to 15 seconds total	Builds rhythm over a longer explanation

If a sequence of b-roll runs longer than about 15 seconds, the viewer starts to forget who is talking. Cut back to the speaker, even for 2 seconds, then return to the images.

Change shots on breaths and commas rather than mid-word, so each cut lands on a beat in the voice.

Why you should not cover every second with b-roll

A popular piece of advice says retention goes up when the picture changes every 2 to 3 seconds, so you should stack b-roll over almost the whole video. I think that advice is borrowed from short vertical clips and applied to the wrong format.

In a 6 to 15 minute talking head video, the speaker is the reason people clicked. If the face is visible for only 20 percent of the runtime, viewers stop building any connection with the person, and the video starts to feel like stock footage with a voiceover. On longer edits I have tested, keeping the speaker on screen for roughly 50 to 70 percent of the time kept comments and repeat views higher than the heavily covered versions.

Constant cutaways also hide pacing problems. If the edit only works when it is wallpapered with images, the interview edit underneath needs tightening first.

Color, motion and performance details that make b-roll fit

B-roll shot on a different day or camera often looks like it belongs to another video. A few fixes go a long way:

Match white balance and exposure to the interview before any creative grade. Put a shared adjustment layer or grade over track 2 so every cutaway shifts together.
Add gentle motion to still images, about 5 to 10 percent scale change over the clip, so photos do not feel frozen next to live footage.
Keep any stabilisation warp below the point where edges start to wobble. A slightly handheld clip looks more natural than an over-processed one.
If 4K b-roll makes playback stutter, generate proxies so you can trim to the frame. The proxy editing workflow for 4K footage explains how to set that up on a modest laptop.

Finally, watch the whole edit once at full screen with fresh eyes, then once with the sound muted. The first pass shows where cutaways interrupt the speaker. The second shows whether the images tell a coherent story on their own.

Frequently asked questions

How much b-roll should a talking head video have?

For videos between 5 and 15 minutes, b-roll covering 30 to 50 percent of the runtime is a sensible starting point. Short vertical videos can carry more because they rely on fast visual changes. Let the content decide, since a heavily visual topic naturally needs more cutaways than an opinion piece.

Can I use stock footage as b-roll?

You can, but use it sparingly and choose clips that show something specific to what is being said. Generic stock shots are easy to spot and weaken trust in the speaker. Footage you shoot yourself, even on a phone, almost always fits the video better.

Should b-roll have its own sound?

Low natural sound from the b-roll, mixed around 20 to 25 dB under the voice, can make a cutaway feel alive. Keep it subtle and fade it in and out over a few frames. If the background sound competes with the speaker at all, mute it.