The Film That Gave the Guide Away: Seven Versions, and the One Change That Worked

In August 2026 we made a short film to promote MH Legal Hub’s guide to the rentista visa, Argentina’s residency route for people with a steady passive income. It took seven versions. Looking back at the rejections, five of the six were about things that could never have saved the film.

The first version ran 20 seconds at 128 beats a minute, with eight cuts in 7.5 seconds, over flat pages of the guide. It performed energy, and it made a free legal guide feel like a trailer. The notes said the music was too intense and the movement too fast. They were right, so we fixed both.

Then the notes changed subject, round after round, and we answered each one properly. We used real photography, a shared colour grade, new music and a written brief, and we ran the whole film through an internal review against our own design charter. Every round made the film better made, and none of them made it good. The fifth version was the most polished yet, and the note on it was one word, repeated: boring.

What the notes were saying

“Boring” and “ugly” are symptoms. People rarely name the defect itself, and a team that repairs symptoms one at a time can do that for weeks.

There is a tell. When the notes change category every round and no fix holds, the problem is above the level anyone is working on. At that point we stop repairing, reread the first message of the project, and ask again what the film is for.

The note on the fifth version carried the answer: show what is inside the guide. The film existed to make someone want the guide, and we had been making a mood. At the end of each version a viewer knew the guide existed and nothing else. It was a teaser, and teasers are the default failure, because a line that hints is much easier to write than a line that answers.

The version that passed changed what the film showed. We storyboarded it before touching the edit. It took real questions from the guide and gave the guide’s real answers, one after the other, with less new polish than the version before it. It was accepted at once.

Giving the content away felt like a risk, and it worked in the opposite direction. A viewer who gets three good answers in under a minute believes the rest of the guide is worth having. A viewer who gets a hint has no reason to believe anything.

Five rounds of craft could not repair a film with nothing to say. One round of content did.

The column that was missing

Every film we make now starts with one written list: what the viewer will know at the last frame. Three to five concrete things, each traceable to a source. If the list will not come, there is no film yet, only a look.

The beat sheet changed with it. A beat column alone lets a teaser through. An answer column does not.

Then we measured

Once the content was right, one honest note remained: some parts carried a lot of text and there was not enough time to read it. That note, and the craft notes around it, could finally be answered with numbers.

Reading time sets the length. Our target is 11 characters a second. The film looked fine and measured 13.5. We took it from 30 seconds to 44, with the same shots in the same order and nothing added. The heaviest line, 59 characters, now holds for 5.4 seconds, which is 10.9 characters a second, and all 27 lines pass. The film’s length came out of that sum. We no longer pick a round number first and pour the text into it.

The safe zone is the platform’s, not the frame’s. A complaint about cut-off text was accurate, and the frame was fine. The text sat where the app’s own buttons and caption lay over the film. The working frame is what remains once the interface has taken its share.

Contrast is measured. One line looked fine on our screens and measured 1.5:1. The accessibility standard asks for at least 4.5:1 for normal text and 3:1 for large text (World Wide Web Consortium, 2018). The eye of someone who already knows what the line says is the worst possible judge of it.