A short social media video that holds attention. Plan it shot by shot.

A good short social media video is planned on paper before anyone presses record. You choose one idea, write a script to a set length and give each sentence its own shot. You also decide where the loop starts and where the thumbnail comes from. Then filming and editing are a matter of working down a page you've already written.

The advice here comes from our own guide, "How to shoot a great short social media video", published by Hynca Media Desk on 4 October 2026. It's a how-to piece. Nothing in it marks a rule change or a new platform feature. Below we take its steps in the order you'd use them on a real shoot, with the sums worked through so you can use them on your next video.

How long should your video be?

Start with one idea per video. That might be a single tip, one product or one change your members need to hear about. If you catch yourself writing "and also", you have two videos. Split them.

The guide sets the ceiling at under three minutes. Within that, it names 30 to 45 seconds as a sweet spot on YouTube Shorts. It gives no evidence for that figure. It also doesn't say whether the advice carries over to TikTok or Instagram Reels. If you post there, treat 30 to 45 seconds as a starting point and let your own results tell you the rest.

Length is far easier to control when you count words. The guide puts most people's speaking pace at 150 to 220 words a minute; about 190 is typical for a short. No source is given for those figures either, but they work as a rule of thumb. At 190 words a minute, a 30-second video is about 95 words of script and a 45-second video about 140. A video just under three minutes long runs to close to 570 words. That's a lot of talking to hold someone's attention through.

Your own pace shifts the sums, and the range is wide. A slow speaker near 150 words a minute fits about 75 words into 30 seconds and about 112 into 45. A fast speaker at 220 words a minute fits about 110 and about 165. So read a paragraph aloud against a timer, once. You'll know which end of the range you sit at, and you'll have the word count for every script after that.

The common mistake is to film a loose, unscripted take and hope to trim it later. Trimming rarely saves it. A 95-word script read at a normal pace gives you a 30-second video with nothing to rescue.

How should you frame and film it?

Film vertical. The guide recommends one loosely framed master at 9:16, 1080x1920, so the same footage can be cut to different lengths. Loose framing means leaving some room around you or your subject. You can then tighten the picture in the edit without cropping into anything that matters. For a small team, that's the main practical gain: one set-up and one recording give you a 30-second cut for one post and a longer cut for another.

Vary the picture with two or three locations and some B-roll. B-roll is the supporting footage that plays while you talk: hands at work, a product on the counter, the room where your members meet. A small number of moves keeps the video lively. It also stops the day turning into packing and unpacking kit.

Plan the locations against the script. If the opening line is about your premises, you might film it outside. If the useful part is a demonstration, film it where the work happens. Each move needs a reason in the words. Without one, it reads as a change of scenery for its own sake.

How do you turn a script into shots?

The guide uses Chris Goor's H.I.D.E. method (from The Studio, Brand Video Pro), which pairs each line of the script with a shot. A wide shot carries the hook. A mid shot carries the intrigue, and a mid close-up carries the value. Then the camera goes back out for the call to action.

The source doesn't spell out what the letters H.I.D.E. stand for. It also doesn't explain what the hook, intrigue and value should contain beyond the shot each one gets. What it does give you is a shape: the camera moves closer as the video builds, then pulls back for the ending. Each change of shot is also a natural place to cut.

The biggest benefit is for people who don't enjoy presenting. With one sentence per shot, you only have to remember one sentence at a time. You record the hook in the wide shot, stop, check it, and move to the mid shot for the next line. Nobody has to deliver 140 words to camera in a single take.

Here's how the page might look for a 45-second video. Write each sentence of your roughly 140 words on its own line. Next to the hook, write "wide". The intrigue lines get "mid", the value lines get "mid close-up", and the final line goes back to "wide" for the call to action. Add the location for each line, plus any B-roll that will play over it.

The page also tests the script. If a line can't be filmed as a single shot, it's usually trying to say too much. Split it, or cut it.

How do you make people keep watching?

Each of the guide's techniques for holding attention belongs on the shot list, next to the line it goes with, before filming starts.

Put the hook on screen in the first frame. The opening line should be visible as text from the very start, as well as spoken. Plan where that text will sit when you frame the wide shot. It mustn't cover your face or the thing you're showing.

Make the video loop by walking into the first shot and out of the last. When the video ends and plays again, the movement joins up and the repeat reads as one continuous piece. This needs the most planning. Decide where you'll walk in from and where you'll walk out to, and make sure the first and last shots are in places that make sense together. If the first shot is in the doorway and the last is at the back of the room, the join won't work. The guide doesn't say how long the loop or the individual shots should be, so judge that from your own edits.

Pick a thumbnail moment while you film. Hold a clean frame for 2 to 3 seconds, with a clear, uncluttered picture. That gives you a frame to choose in the edit, so you aren't stuck with a blurred still caught mid-word. The end of the mid close-up, once the value line is delivered, is often a sensible place for it.

How do captions and keywords help people find it?

Burn in word-by-word captions. The guide's reason is that many people watch with the sound off. Captions fixed into the picture mean the video still makes sense to them. Word-by-word captions follow your speech closely, so viewers can keep up without reading ahead. Check each caption against the audio before you export. A misheard word on screen is easy to miss and hard to unsee.

Choose one keyword for the video before you write. Then say it aloud and show it on screen. According to the guide, platforms use speech to understand what a video is about, so the spoken word and the on-screen text both tell the platform what your short social media video covers. The guide gives no evidence for that claim. Saying the keyword costs you nothing, though.

Put the keyword where it fits naturally, usually in the hook or the value line. If you have to force it in, the sentence will sound forced to viewers too. Once it's written into a line on your shot list, it won't be forgotten on the day.

See how the Desk drafts video scripts from your source facts, held for your approval. Nothing is published by itself.