The 4-Part Reel Script Structure That Actually Converts
Hook, problem, value, CTA — why this order works, what each beat is responsible for, and how to write all four in under a minute.
Summary — the short answer
- A short-form script has four beats — hook, problem, value, CTA — and each has one job.
- Write them out of order: value first, then problem, then hook, then CTA.
- The hook is a promise; the value beat is the payment. Writing the hook first is how creators end up over-promising.
- Budget roughly 0–3s hook, 3–8s problem, 8–18s value, 18–22s CTA for a 22-second video.
- Ask for exactly one action. Follow, save, comment and click in the same CTA produces none of them.
Key facts
- Beats
- Hook, Problem, Value, CTA
- Writing order
- Value → Problem → Hook → CTA
- Ideal length
- 90–130 words for 22 seconds
- Hindi adjustment
- About 15% fewer words
- CTA rule
- One action only
A short-form script is not a shorter long-form script. It is four decisions delivered in the order a stranger needs them, and the order is not negotiable. Change it and you lose people even when the individual sentences are good.
The four beats and their jobs
| Beat | Timing | Its one job |
|---|---|---|
| Hook | 0–3s | Make a specific promise |
| Problem | 3–8s | Name the objection already in their head |
| Value | 8–18s | Deliver exactly what you promised |
| CTA | 18–22s | Ask for one easy action |
Hook
The hook is a promise small enough to be believable and specific enough to be wanted. "This changed everything" is neither. "Maine 30 din tak yeh serum use kiya — result honestly shocking tha" is both: there is a defined period, a defined subject, and an implied payoff.
Problem
The problem beat is not background. It is the objection the viewer is already holding, said out loud so they feel understood before you have proven anything. If your problem beat could be deleted without the video making less sense, it was context, not a problem.
Value
This is the payment for the promise. It is the only beat that is allowed to take real time, and it should be concrete: a number, a step, a specific product, a mechanism. Vague value beats are why videos die at fifteen seconds — the viewer stayed for the payoff and it never arrived.
CTA
One action. The easiest possible one that still serves you. Asking for four things at once is the same as asking for nothing, because the viewer has to make a decision and the default decision is to scroll.
Why you write them out of order
Start with the value beat. Decide precisely what the viewer will know or be able to do at the end of the video. Then write the problem as the objection that makes that value matter. Only then write the hook, as the shortest possible promise of that value.
Creators who write the hook first almost always promise something the video never delivers, because the hook was written before the payoff existed. The retention graph shows the exact second the viewer noticed the gap, and it is usually around eight seconds.
A complete worked example
- Value: three drops of a ceramide and niacinamide serum at night repairs a damaged skin barrier in about four weeks.
- Problem: dull skin, a new product every week, nothing working, and a suspicion that the whole category is a scam.
- Hook: "Maine 30 din tak yeh serum use kiya — aur result honestly shocking tha."
- CTA: "Link bio mein hai — apna before/after 30 din baad comment karna."
Notice that the CTA asks for a comment thirty days later, which is a low-friction action that also creates a second touchpoint. That is a CTA doing two jobs without asking for two things.
Length, language and pace
For a 22-second video, aim for 90 to 130 words in English. Hindi generally needs about 15 percent fewer words for the same idea, because spoken Hindi carries more syllables per concept. Hinglish sits between the two and is the most forgiving, which is one reason it dominates Indian short-form.
Testing the structure
Hold the problem, value and CTA constant and vary only the hook. That isolates the variable that explains most of the performance difference between two videos. Once you have a hook that works, reuse its pattern rather than the exact line — patterns generalise, lines do not.
Frequently asked questions
Does this structure work for entertainment content, not just education?
Yes, but the beats compress. The hook is often the first frame of action, the problem is implied by setup, and the value is the punchline. The order still holds.
How long should each beat be for a 60-second video?
Roughly double everything except the hook. The hook stays under three seconds regardless of video length, because the decision to keep watching happens at the same speed.
Can I skip the CTA?
You can, and for pure reach plays it sometimes helps. But a video with no ask converts viewers into nothing, so use silence deliberately rather than by default.
Should the hook appear as on-screen text too?
Yes. Most short-form is watched muted at first, so the hook needs to work as text as well as audio. Keep the on-screen version shorter than the spoken one.
Sources and further reading
- Instagram Creators — making ReelsPlatform guidance on structure, length and what Reels distribution rewards.
- YouTube — Shorts best practicesOfficial documentation on Shorts format and creation.
- VerbCraft: hooks that survive 1.5 secondsA deeper look at the hardest beat to write.
- VerbCraft: brand voice for AI scriptsHow to make generated scripts sound like you rather than like everyone.
- VerbCraft pricingUnlimited script generation with ranked hooks is included on the Creator plan.
Use this article elsewhere
Copy a structured brief for ChatGPT, Claude, Perplexity or Gemini — it includes the key points and the canonical link so the assistant can cite VerbCraft properly.
https://verbcrafts.in/blog/four-part-reel-script-structure
