The 90-Second Rule for Video Scripts

Most short-form video scripts die at the same spot.

The retention dashboard across thousands of clips shows the same shape. Watch time falls off a cliff right at the 30-second mark.

The drop is structural. Same place every time, across creators, niches, and content types.

Most scripts run setup-setup-setup-payoff-CTA. By the time the payoff lands, the audience has already left.

The 30-second cliff is the algorithm's decision point. Past it, reach is capped at the audience you already have. Before it, you get amplification.

The 90-second rule for video scripts is the structure that survives the cliff. Three beats inside 90 seconds, hook in the first five, payoff in the back two-thirds.

90 seconds is the cap on what counts as short-form. Inside that 90 seconds, the script structure decides whether the algorithm feeds it.

Key Takeaways

• Short-form retention falls off a cliff at 30 seconds across most clips. The fix is putting the payoff hook inside the first 15 seconds so the viewer sees the answer before the cliff arrives.
• 3 beats: hook, tension, payoff. Adding beats past 3 dilutes the payoff because every extra beat steals time from the one that matters.
• One third of the runtime is hook and setup. Two thirds is payoff. Inverted ratios produce cliff-shaped retention curves.
• The 5-second summary test: at second 5, can a viewer summarize the payoff in one sentence? If yes, the next 85 seconds are earned. If no, the algorithm stops feeding it before second 30.
• 90 seconds is the working cap. Past 90 the audience expectation flips and the retention curve looks completely different.

The 30-Second Cliff

Short-form algorithms decide amplification off the first 30 seconds of retention. The cliff is where the watch-time curve falls. The cap is the reach decision the algorithm makes when the cliff is steep.

I've watched dashboards where retention falls off right at the 30-second mark across thousands of clips. The drop is structural. Same place every time.

Most scripts lose people at the same spot because the structure doesn't escalate. The hook does its work in second 5, then the next 25 seconds run setup the viewer already paid for with the hook. The payoff lands at second 45. By then the audience left at second 28.

The 30-second cliff is about pacing. Content quality matters too, and bad pacing kills good content before quality has a chance.

What clears the cliff is putting the payoff hook inside the first 15 seconds. Tease the answer, then explain. The viewer sees the payoff signal early, decides the next 60 seconds is worth their time, and the retention curve flattens past 30.

Pushing scripts to 90 seconds or 2 minutes leaves the cliff at 30 seconds. The cliff is structural. Pacing is what gets the viewer across it.

I've watched creators move from 60-second scripts with weak pacing to 45-second scripts with sharp pacing and triple their watch time. The shorter script worked because it respected the cliff. The longer script collapsed because it ignored the cliff.

The 3 Beats

Hook, tension, payoff. Three beats, every short.

The hook earns the next 5 seconds. The tension keeps the viewer through the middle. The payoff is the reason they watched. Anything else is decoration.

I've seen scripts try to fit 5 beats into a 60-second clip and lose the audience at beat 3 every time. Adding beats dilutes the payoff. Every extra beat steals time from the one that matters.

3 beats is the cap. If a script has 4, one of them has to come out before filming. If a script has 5, two come out.

The discipline of cutting to 3 beats forces the script to surface what's load-bearing. The opening line, the tension that holds the middle, and the payoff that resolves the loop. Everything else lands on the cutting room floor.

A 3-beat script with a tight hook and a sharp payoff outperforms a 5-beat script with average pacing every time. The audience can't follow 5 beats inside 60 seconds. The audience can follow 3.

The 3-beat structure also makes editing trivial. The cuts are clean because the beats are discrete. Multi-beat scripts blur the cuts and the editing gets harder, which is the symptom of a structural problem upstream.

The 3 beats are the structural floor. Every short worth shipping has them, even if the creator can't name them.

The Hook-to-Payoff Ratio

What ratio of your script should be hook and setup versus payoff?

Hook and setup get one-third of the runtime. The payoff gets two-thirds.

In a 60-second script, that's 20 seconds of setup, 40 seconds of payoff. In a 90-second script, that's 30 and 60. The ratio scales with the total length, but the proportions stay fixed.

I've audited scripts where the hook ran 45 seconds and the payoff got 15. Inverted ratio. The retention curve looks like a cliff because the audience left while the script was still setting up.

The payoff is the part where the viewer learns the thing they came for. Sometimes that's a big reveal. Most of the time it's the specific tactic stated cleanly, with enough context to make it usable.

The hook earns attention. The payoff earns the follow.

Inverted ratios are the most common script error I see. The script-writer wants to set up the payoff carefully, so the setup eats half the runtime. The payoff gets compressed into the back third, and the punchline lands in a script that already lost the audience.

The fix is timing the first draft. Write the script, then read it aloud with a stopwatch. If the hook plus setup runs past one-third of the target length, cut. Move setup into the payoff section if needed. The audience can absorb context inside the payoff. Context after the audience has scrolled lands nowhere.

The Watched-vs-Scrolled Test

There's one test that tells you whether your script gets watched or scrolled.

The 5-second summary test. At second 5 of your script, can a viewer summarize the payoff in one sentence?

If yes, you've earned the next 85 seconds. If no, the algorithm stops feeding it before second 30.

I've stress-tested this on dozens of scripts before filming. The ones that fail the 5-second test fail in production every time. The ones that pass have a chance at the algorithm.

The test catches the failure mode where the script writer thinks the payoff is implied. The viewer needs the payoff signal explicit in the first 5 seconds, even if the answer itself lands later. Implied payoff signals get scrolled past.

The fix is restructuring so the payoff signal lands inside the first 5 seconds. The setup runs in the back half of the script, after the audience already knows where you're going.

This is the inversion most scripts need. Setup becomes throat-clearing in the new structure. The payoff signal opens. The setup supports it, after the audience has committed.

Once the test is internalized, it changes how scripts get written. The opening line carries the payoff signal directly. The script earns attention by promising the answer in the first 5 seconds, then delivering it in the next 30.

The 90-Second Cap

90 seconds is the cap on short-form. Past 90 the algorithm changes how it treats the clip.

YouTube Shorts caps at 60. TikTok caps at 10 minutes, and the algorithm rewards under 90. LinkedIn video reposts go further.

The 90-second working ceiling is what matters across the mix.

Past 90 seconds, the audience expectation flips from "stop the scroll" to "watch a thing." The retention curve looks completely different. The viewer commits longer up front, but the cliff at 30 seconds gets steeper because the algorithm starts comparing the clip to long-form content for retention benchmarks.

I've watched the same content perform 4x better at 75 seconds than at 110 seconds. Same hook, same payoff, longer middle. The longer middle cost reach.

If a script is running long, the cut goes in the middle. Tighten the setup or compress the explanation. The hook and the payoff stay. Setup and explanation are where 90% of cuttable time lives.

The 90-second cap is a discovery about how short-form algorithms behave. Working with it is easier than working against it.

There's a related constraint at 45 seconds. The retention curve sometimes shows a second drop around 45 to 60 seconds for content that doesn't escalate the payoff fast enough. A 45-second script with sharp pacing often performs better than a 90-second script with the same content stretched out.

Pacing per second matters more than absolute length. A 30-second script can hold an audience if every second carries weight. A 90-second script that pads loses them at second 12.

What this changes about your weekly video output

If the team's video output runs to a few 2-minute clips per week, the 90-second rule says cut them in half. If the output runs to one 60-second clip per week, the rule says ship two. The shorter format multiplies output without dropping quality.

The structural discipline of the 3 beats also speeds up writing. A 3-beat script with the right ratio takes 20 minutes to draft. A 5-beat script with inverted pacing takes 90 minutes and still doesn't work.

Next week we'll cover the H1 attribution post-mortem: the 3 questions to ask about every channel before H2 planning, and the one spreadsheet that kills the ghost channels on your report.

For now, the 90-second rule fits on a sticky note. Cliff at 30. Three beats. One-third hook, two-thirds payoff. 90-second cap.

The script that respects all four publishes Monday.

Further Reading

On Professor Leads

• Hooks That Stop the Scroll is the upstream hook-pattern framework that feeds into the 3 beats covered here.
• AI Creative Without the Slop covers the creative-workflow side of the same video production system.
• The Psychology of Lead Nurturing is the audience-attention framework that the 5-second summary test draws from.

On Forbes (by William DeCourcy)

• Beyond The Click: Why Lead Generation Culture Must Evolve Now is the bigger argument about audience attention and how content has to evolve to earn it.
• William DeCourcy on the Forbes Business Development Council for related Council writing on marketing, lead generation, and team building.