Short-form video has a brutal honesty to it. The platform's view metric kicks in at three seconds, the algorithm rewards anything that holds beyond that, and the audience has a thumb hovering over the screen the entire time. Almost everything you can learn about whether a piece of creative will work, you learn in those opening frames.
That doesn't mean the rest of the video stops mattering. It means we plan it differently.
Briefing for the opening
Before a shoot, we write a separate one-line brief for the first three seconds of every concept. It's a tiny artefact, but it forces the team to articulate what the viewer is supposed to see, hear, and feel before they've made the decision to keep watching. If we can't write that line cleanly, we don't have the concept yet, no matter how good the rest of the script looks.
The brief usually sounds like a question, an unexpected image, or a contradiction the viewer can't immediately resolve. What it isn't, is the brand logo, the product hero shot, or the studio identifier. Those have their place. Their place is not the first frame.
Shooting for the opening
On set, we shoot the opening multiple times in multiple configurations. Different framings, different energy, sometimes radically different scripts for the same idea. The cost on the day is marginal, the value in the edit is enormous. The best opening is rarely the one the storyboard called for.
We also shoot in a way that lets us re-cut. Wide and tight on every important moment, audio captured clean and separately, and reactions held a beat longer than you think they need. The third frame might end up being the second take of the sixth setup. We want it to exist before we know we need it.
Editing for the opening
In post, every cut starts from the opening, not the storyboard. The editor picks the strongest three seconds first and works outward from there. If the opening is honest, the body of the spot has to live up to it, not introduce it. Most ads do the opposite: they spend the first five seconds setting up a payoff that never lands because nobody watched long enough to see it.
We also test the opening on its own. A three-second cut, no resolution, no offer, posted as a piece of organic content or shown to a small paid audience. If a viewer leaves a comment or sticks around long enough for the platform to register, we know we have something. If not, we don't waste a finished edit on it.
The rest of the spot
None of this is an argument for ignoring the second half of a video. It's an argument for earning it. The audience pays attention in instalments. The first three seconds buys you the next ten. The next ten buys you the offer. The offer is what we built this for, and we'd like more people to actually see it.