We started making six second cuts of our thirty second films, because that was what the placements wanted and it seemed like an editing task.
They performed badly for a year. What eventually worked required starting from scratch, and the difference is structural rather than a matter of length.
What does not fit in six seconds
The obvious answer is most things, and it is worth being specific about which.
Any narrative arc. Setup, development and resolution cannot happen. Attempting it produces something that feels like a trailer for an idea rather than an idea.
Any argument. Establishing a problem and presenting a solution needs more time than this.
Any brand reveal at the end, which is how thirty second films are conventionally structured and which in six seconds means the brand appears just as attention is leaving.
And anything requiring a viewer to read more than a few words.
What does fit
One thing. Genuinely one, which was the finding that changed our approach.
A single visual idea, understood immediately, with the brand present throughout rather than at the end.
A single line, said or shown, which is the whole message.
A demonstration of one property of a product, shown rather than described.
Or a single joke with a fast setup and a fast payoff, which is the hardest to do well and the most effective when it works.
The successful formats we landed on all shared the property that a viewer could describe what they had seen in one sentence.
The first frame does most of the work
The technical finding that mattered most.
In a six second unit, the opening frame is a substantial proportion of the total exposure. A slow fade in, or a logo card, or an establishing shot, consumes a fifth of the available time before anything happens.
Everything we made improved when we cut the first second entirely and started at the moment of interest.
The same applies to the last frame. An end card held for a second and a half is a quarter of the ad. We moved to having the brand present throughout, small and constant, rather than as a terminal card.
Sound cannot be assumed
The constraint that shapes everything else in most placements.
A large proportion of these units are served without sound initially, and in six seconds there is no time for somebody to decide to turn it on.
So the piece has to work silently, which means the meaning has to be visual, and any words have to be on screen.
Where sound is used, it works best as reinforcement of something already visually clear rather than as the carrier of the message.
Text on screen has its own constraint. Reading takes time, and a viewer can read perhaps five or six words comfortably in the available window while also looking at an image. Longer copy is not read at all.
What the measurement told us
Worth noting because it shaped what we optimised for.
Completion rate is close to meaningless on very short units, because most of them complete regardless of whether anybody was paying attention.
What differentiated our better performers was brand recall in survey testing, which required commissioning proper brand lift studies rather than reading platform metrics.
The pieces that scored well on recall were the simplest ones, consistently. The ones with more content in them scored worse, which is the opposite of what the team's instinct produced.
How we make them now
Written as six second pieces from the start, never cut down.
One idea, decided before anything is shot, and written as a single sentence that everybody agrees on.
Brand present from frame one.
No fade in, no end card, no dialogue.
And produced in sets, because they are cheap to make once the idea exists and because variation within a campaign performs better than repetition of one execution.
The relationship with the long form
The thing we resolved last.
We stopped treating short units as a version of the campaign and started treating them as a separate expression of the same idea.
The thirty second film tells the story. The six second units each take one element of it and state it plainly.
They share a look and a line and a brand, and no footage. That costs more than cutting down and it is the only approach that has worked for us.
Production economics
A commercial note, since the format changes the budget shape.
Short units are cheap to produce individually and the cost sits almost entirely in the idea rather than the shoot.
Which argues for shooting a set in one session, from one setup, rather than treating each as a project.
We now produce eight to twelve variants in a day for roughly what one thirty second film costs, and the variation itself improves performance because frequency capping is imperfect and repetition of a single execution wears out fast.