ShortForm Studio

Workflow · Stage 1 of 5

Plan — pre-production calculators

Five tools you run before you press record. Each one turns a fuzzy judgement ("is this hook strong enough?") into a number you can defend in a debrief.

Tool 1 · Hook Score Analyzer

Score the first 3 seconds of your script.

Paste what you plan to say out loud in the opening 3 seconds. We score 7 criteria and tell you whether the hook earns the next 7 seconds.

Time: ~2 minutes

Result appears here. Past a strong hook first — the better the input, the more useful the score.

Why this matters

On short-form platforms, the first 3 seconds decide whether you get the next 30. Most creators lose viewers not because the video is bad, but because the opening spends its first frame on a logo, a wave, or a half-sentence of context. The 7 criteria below are the levers you can actually move in a rewrite.

Limitations

  • It scores the line itself, not the visual that runs over it. A weak line can still work with a strong visual, and vice versa.
  • It cannot detect delivery, voice tone or facial expression — which are often half the hook.
  • LinkedIn rewards authority; TikTok rewards curiosity. Different platforms weight the same words differently.
Frequently asked questions

What counts as an emotional trigger word? Anything that maps to a feeling — "secret", "mistake", "regret", "relief", "tonight", "free", "banned". The analyzer uses a 90-word list. If your word isn't on it, the score stays neutral.

Why does the platform change the score? LinkedIn values authority hooks ("After 10 years of editing…"); TikTok values curiosity gaps ("Nobody tells you this about…"). Same words, different weights.

Can I score a visual-only hook? Yes — leave the text field blank and pick "Pure watch time". You'll get 6 of 7 criteria; the "line" criterion is skipped.

Tool 2 · Script Pace & Word Counter

How long is your script, and how fast does it have to be?

Pastes a full script. Returns word count, speaking duration, pace classification per platform, filler-word count and longest sentence.

Time: ~1 minute

Paste a script and pick a delivery style to see the breakdown.

How to read the output

The pace band tells you whether your words per minute fall inside the comfortable range for the platform you chose. Below 120 wpm and viewers feel you are dragging; above 200 wpm and they retain less. The "filler density" score is the percentage of words that are typical fillers ("um", "like", "you know", "basically"). Anything above 4% is where viewers tune out.

Limitations

  • The tool assumes English. Other languages will report inaccurate durations.
  • It cannot measure pauses you insert for effect — those are valuable and should be kept.
  • Numbers assume straight-to-camera talking. If you do heavy voiceover over B-roll, multiply by 0.85.
FAQs

Why does my pace read "too fast" but my video feels normal? Probably because you paused for visuals. Re-time yourself on a single take with no cuts, then compare.

What's a good filler density? Under 2% is excellent. 2–4% is normal for unscripted talking. Above 4%, consider a script rewrite.

Tool 3 · Storyboard Panel Calculator

How many storyboard panels does your video actually need?

Most "30-second" storyboards only have 4 frames — which is why edits feel choppy. This tool estimates a realistic panel count based on your scene count and average shot length.

Time: ~1 minute

Set your numbers and the panel recommendation + per-scene breakdown appears here.

How to read this

We subtract transition time, account for an opening title card (1.5 s) and closing card (1 s), then divide what is left by your average shot length. The number is the realistic minimum panel count. Drawing fewer panels forces you to fake coverage in the edit, which is the leading cause of "this cut feels too long".

Limitations

  • It assumes one shot per panel. Real coverage usually needs 1.5× this many.
  • Talking-head videos with one continuous take don't need panels — set scenes to 1.
  • Animation / motion design is harder to estimate; multiply by 1.8×.
FAQs

Why does MrBeast-style hyper-cut need so many panels? His average shot length is around 1.8 s. A 60 s video at that pace is ~33 cuts, each ideally planned.

Can I draw thumbnails instead of every panel? Yes — but only for films you're confident in. For first cuts, panels are cheaper than reshoots.

Tool 4 · B-roll Shot List Generator

A structured B-roll shot list, generated from your topic.

Enter the topic of your video and a few constraints. Get 10 specific shot suggestions, organised by purpose (establishing, detail, transition, emotional), each with a recommended shot type and estimated shoot time.

Time: ~2 minutes

Your shot list appears here. Each item shows shot type, framing and why it earns its place.

How the list is built

We pair the topic against a curated map of B-roll archetypes (process, texture, context, reaction, transition). For "pour-over coffee in a calm setting", a process shot of the kettle is almost always needed; a top-down texture shot of the beans earns its place because coffee content lives or dies on detail. The list is deterministic — same inputs always produce the same list.

Limitations

  • It cannot see your space. Adapt each shot to what you actually have.
  • Some topics need specific props we cannot know about (a specific product, person).
  • It suggests; it does not direct. A great B-roll list still needs your taste.
FAQs

Why deterministic instead of AI-generated? Because the same topic in the same setting should always produce the same scaffolding. AI gives novelty; we give consistency. You bring the novelty.

Can I edit the list? Treat it as a draft. Re-order, delete, add — but keep one shot per category so coverage stays balanced.

Tool 5 · Intro Drop-off Predictor

Predict where viewers leave — before you publish.

Inputs hook score (from Tool 1), intro length, and first visual change. Outputs a 3-point drop-off curve and a "first visual action" gap that is the most common early-loss cause.

Time: ~2 minutes

Set your numbers; the predicted retention curve appears here.

How the model is built

A simple weighted blend of public short-form retention studies (Stanford, Tubular Labs creator reports) plus industry rules of thumb. The largest drop is always at the first visual action delay + 2 seconds; a second drop appears around the intro / payoff transition; a third at the video's length-vs-platform median (Reels ~17 s, TikTok ~22 s, Shorts ~32 s).

Limitations

  • It uses average audience behaviour; your niche may retain longer (tutorials) or shorter (memes).
  • It does not model replays or loops — videos people rewatch show different curves.
  • It cannot detect trending audio lift, which can flatten drop-off for a few days.
FAQs

What is "first visual action delay"? The gap between the very first frame and the first clear visual change — a cut, a movement, a text pop. Under 1 second is excellent. Over 2 seconds is where most early loss begins.

Why does hook score matter more than intro length? A great hook buys you time. A weak hook even with zero intro will still lose viewers because there is nothing to follow.

Next stage: Shoot →

Advertisement (Natural placement below content)