Brand Film Structure: A Working 3-Act Blueprint for 60–120 Seconds

A 90-second brand film is not a small feature. It is closer to a pop song: a fixed runtime, a known shape, and no room for a scene that merely exists. The brand films that fall apart do it in the outline stage — someone shoots a day of nice material with no beat map, and the editor is left arranging pretty shots into a shrug. The ones that work were structured before anyone touched a camera.

This is the structure I use for 60 to 120 second pieces, whether the client is a bakery or a B2B platform. Four beats, ruthless scene budgeting, and a shot plan that answers to the beats instead of to whatever looked good on the location scout.

The four beats: hook, tension, proof, resolve

Call it three acts with the second act split in half. Every effective short brand film I have cut or studied lands on some version of this shape:

1. Hook (roughly the first 5–10 seconds). A striking image, a bold claim, or a question that names the viewer's problem. Not the logo, not an establishing drone shot. The film earns its runtime here or nowhere.
2. Tension / turn (the next 20–35 seconds). The world before, or the problem at full weight. What was hard, broken, slow, or at stake. Then the turn: the moment the brand, product, or founder decision enters. Tension is what makes the brand's entrance feel like relief instead of an ad.
3. Proof (the biggest block, often 30–50 seconds). Show, do not assert. The product doing the thing, the process in real hands, the customer whose day looks different, the specific detail that only this brand would know. Proof is where b-roll actually belongs.
4. Resolve (the final 10–20 seconds). Emotional landing plus one clear ask. Where things stand now, said in one line, and a single call to action. One. A resolve with three CTAs resolves nothing.

Scene budgeting: spend seconds like money

Before writing a script or shot list, budget the runtime. A 90-second film has 90 seconds — write the number for each beat at the top of the outline and treat overruns like budget overruns, because they are. My default split for 90 seconds: hook 8, tension and turn 27, proof 40, resolve 15. For a 60-second cut, compress tension hardest; for 120, expand proof, never the resolve.

Then budget scenes against those seconds. A scene — one location, one activity, one idea — needs about 8 to 15 seconds of screen time to register and breathe. That means a 90-second film holds roughly six to nine scenes, total. Every additional location on the shoot day that does not map to a budgeted scene is a location you are shooting for the trash bin. This single discipline shortens shoot days, cuts costs, and makes the edit dramatically faster — the same principle behind editing videos faster: decisions made early are edits you never have to make.

Interview-driven vs scripted VO

The spine of the audio comes from one of two places, and the choice changes how you shoot the entire film.

Interview-driven films build the narration from a real conversation with a founder, maker, or customer. They sound human because they are, and imperfect phrasing reads as sincerity. The cost is control: you must interview for the beats, asking questions that produce a hook line, a tension line, and a resolve line — then log the transcript against the beat map before you plan b-roll. Shoot the interview first whenever the schedule allows, so the b-roll list serves lines you actually captured instead of lines you hoped for. Audio quality is non-negotiable here since the voice carries the film.

Scripted VO films trade warmth for precision. Every word is chosen, the runtime is known before the shoot, and the picture can be boarded exactly. This suits product-led films, films with heavy legal review, and brands whose founders freeze on camera. The risk is the plague of brand films that sound like a press release read aloud — fight it by writing the VO in the second person, short sentences, contractions, and reading every draft out loud before anyone approves it.

A practical hybrid: interview for hook, tension, and resolve — the emotional beats — and a few lines of tight VO or on-screen text for the proof section, where precision about what the product does matters most.

The shot plan, mapped to beats

Here is the planning table I actually bring to set for a 90-second interview-driven piece. Yours will differ in content, never in shape: every row ties a beat to seconds, to specific coverage, and to the feeling that coverage must produce.

Notice the proof row says sequences, not shots. A wide, a medium, and a detail of the same action cut together into a satisfying 10-second unit; three unrelated singles do not. Shoot proof in sequences and the edit assembles itself.

Common failure modes and the fixes

• Logo-first opening. The brand reveal belongs at the turn or the resolve. Opening on a logo converts the hook into an ad label.
• Proof without tension. Reads as a product tour. Restore the problem beat, even if it is only two lines.
• Interview soup. Six speakers in 90 seconds means nobody registers. One voice, two maximum.
• Multiple CTAs. Visit the site, follow the socials, use the code — pick one. Splitting the ask splits the result.
• Coverage-driven structure. If the outline was reverse-engineered from footage, it shows. Structure first, shoot second, always.

Structure is also a business tool: a beat map with second budgets is a scoping document. It defines what the shoot day must produce, which is exactly the clarity that protects you when pricing the job — a topic covered properly in how much to charge as a videographer. A film that is structured before it is shot is cheaper to make, faster to cut, and far more likely to do the one thing brand films exist to do: leave the viewer with a single feeling and a single next step.