A short-form video system that compounds

Short-form video is a hook contest. If the first two seconds do not earn the third, the rest of the work never gets watched.

A hook is not a trick, it is a promise. Name the outcome, show the result, ask the exact question your customer asks, or open on the most interesting moment of the story.

Write ten hooks for every video, read them aloud, and keep the one that sounds like a person talking. This is the cheapest edit you will ever make.

Read the hook without the video. If it only makes sense once the footage starts, it is not a hook yet.

Formats beat ideas. A format is a recipe: a fixed structure you refill with new material every week, so the team knows what to shoot and the audience knows what to expect.

Pick three. One that answers questions, one that shows process or product, and one that carries the brand's point of view. Rotate them on a schedule.

Trends are seasoning, not the meal. Chasing a viral sound every week produces content nobody remembers and a team that burns out.

Name the formats and write them down. A recipe that only exists in someone's head leaves when that person does.

Filming when the mood strikes is how content systems die. Treat it like a production: one day, one shot list, ten to twenty assets, then back to work.

Build the shot list from the month's calendar: launches, questions from sales, a seasonal moment, one behind-the-scenes idea. Everything else is B-roll, and you can never have too much.

Consistency shows. When the set and the sound are stable, the audience notices the idea instead of the production.

Change one variable at a time: the set, the format, the presenter. Otherwise the data tells you nothing about why something worked.

One shoot should produce a month of content. Shoot vertical by default, but frame with room to crop, and record clean audio separately so you can re-cut without the room tone changing.

From a single session you can cut 9:16 for reels, 1:1 and 4:5 for feeds, 16:9 for YouTube and the site, stills for the newsletter, and three or four paid variants. Shoot once, cut the shapes you know you will need, and archive the rest.

Burn in captions. Most viewing happens with sound off, and captions also help the platforms understand the content.

Publish from one calendar so organic and paid do not compete with the same asset on the same day. A simple weekly slot for each format keeps the machine running without a meeting.

Likes are a vanity metric with a lag. The numbers that change decisions are retention and intent.

Watch where people drop off, which hooks hold, which posts generate saves and sends, and which produce profile visits and messages. Those are the signals that lead to revenue.

Review monthly, not daily. One strong format is worth more than a week of noise.

One more signal is worth keeping: which videos people mention in real life. A post that comes up in a sales call is doing more than its view count suggests.

Organic short-form is the cheapest creative test you will run. The hooks that hold attention unpaid are the hooks to put budget behind.

For a brand where we run paid media, the same footage has to work in the feed and in the ad account: one capture, organic cuts first, then paid variants of the winners. That is the logic behind the campaigns we run for Passion d'Ames.

Then refresh. Creative fatigues, so the system should produce replacements before performance drops, not after.

Keep the winners in a library, labelled by hook and format. Next quarter's creative then starts from evidence instead of a blank page.