The five-beat structure of a UGC script that converts

A converting UGC script has five beats — hook, problem, proof, payoff, CTA — sharing a budget of about 70 spoken words for a 30-second cut. Each beat has one job; the budget stops any one of them from eating the video.

Here's the 30-second version, worked through with a stain remover:

  • Hook — 0:00–0:03, ~8 words. Name the pain or open a loop. "I almost threw this shirt away — watch this."
  • Problem — 0:03–0:08, ~12 words. What the old way costs. "Set-in coffee stains laughed at every spray I'd already tried."
  • Proof — 0:08–0:20, ~28 words. The demo moment: the product visibly doing the thing while you narrate over the footage. "You work it in dry, wait sixty seconds, rinse from the back. No scrubbing — you can literally watch the stain lift out of the fibers."
  • Payoff — 0:20–0:26, ~14 words. The after state, small and specific. "Two washes later and it still hasn't come back."
  • CTA — 0:26–0:30, ~8 words. One action. "It's linked below if you've got a shirt to save."

Notice what's missing: no "hey guys," no brand backstory, no feature list. The hook is the first frame. If you have to cut words, cut anywhere but proof — the demo is the beat that sells.

How many words fit in a 15-, 30-, and 60-second video?

At a natural on-camera pace — roughly 2.2 to 2.5 words per second — you get 32–38 words in 15 seconds, 65–75 in 30, and 130–150 in 60.

  • 15 seconds: hook, one proof line, CTA. No room for a problem beat; the visual has to carry it.
  • 30 seconds: the full five beats. The workhorse length for ad creatives.
  • 60 seconds: not the 30 padded out — usually a story-shaped problem, two proof moments, and a fuller payoff.

Most beginners write 180–220 words for a 30-second brief, then either sprint the read or drift to fifty seconds. Write to the low end of the budget instead. Pauses count as words — one silent second of the stain actually lifting is worth more than any adjective you'd say over it.

Before you film, run the read-aloud test: stand up and read the script out loud, on a timer, at the pace you'd tell a friend a story. Two things happen. You learn your real runtime — usually 40 percent over. And your mouth catches the "written" language your eyes approved: "utilize," "additionally," "boasts," "formulated with." If you stumble on a line, the line is wrong, not you — rewrite it as whatever you accidentally said instead.

Should you memorize, read, or improvise your script?

None of the three, purely. Memorize the hook and the CTA word for word, turn the middle beats into short bullets, and film one beat at a time.

Each pure version fails in its own way:

  • Full memorization goes flat — viewers can hear reciting — and one flubbed word at second 24 restarts the entire take.
  • Reading off notes or a prompter shows in your eyes; the left-to-right scan is visible on camera.
  • Pure improv rambles. Take 14 still doesn't contain the claim the brand needs.

The hybrid works because of how UGC is cut: jump cuts are native to the format, so film each beat as its own take. That's why a script halves your takes — "done" is defined per beat. Nail the hook in three tries and never say it again. Flub the payoff, re-film six seconds, not thirty.

The setup is low-tech: the current beat in huge text in your notes app, propped under the lens. Glance, look up, deliver. If a line won't come out clean after four tries, the line is broken — rewrite it. Framing and light are a separate job — covered in how to film UGC on your phone.

How do you script around a brand's required talking points?

Give each mandatory talking point its own beat and translate it into how you'd say it out loud. Never paste the brief's bullets into the script in a row — three claims back to back is the press-release moment, and "feels scripted" is the revision note it earns.

Map points to beats:

  • How-it-works claims ("dissolves stains at the fiber level") go in the proof beat, said while the demo shows it.
  • Results claims ("gone in one wash") go in the payoff.
  • The offer or discount goes in the CTA — and nowhere else.

Then translate. The brief says "clinically proven to reduce frizz for 72 hours." A person says: "it's clinically tested — and my hair genuinely stayed smooth for three days." Same claim, ear language.

One exception: some phrases are legally locked. Before you write, ask the brand — "which of these have to be said word for word?" Keep locked phrases exact but build your own sentence around them: lead in naturally, land the phrase, move on. And never ad-lib claims the brief doesn't support — especially results or health claims. That's how videos get pulled.

How do you write a CTA that doesn't feel cringe?

Cringe is mismatch — a hard-sell line stapled onto a soft, native-feeling video. The fix is one small, specific action, phrased the way you'd tell a friend.

  • Point, don't push. In a paid placement, the button and caption do the selling. Your line just points: "it's linked below."
  • Specific beats generic. "Check if the three-pack's still on sale" outperforms "don't miss out!"
  • Qualify the viewer. "If your whites are as wrecked as mine were, it's worth a look" filters for buyers without begging everyone else.
  • One ask. Follow-and-click-and-comment is zero asks.

Skip manufactured urgency. "Selling out fast" when you've never seen their inventory reads scam-adjacent, and careful brands cut it anyway. If the brief has a real deadline, say it plainly.

The whole craft compresses into a ritual you can run tonight: write the five beats, read them aloud on a timer, cut to the low end of the budget, film beat by beat. That's the difference between rambling at a lens and delivering something a brand can run.

FAQ

Do brands give you a script, or do you write it? Both happen. Most briefs hand you talking points and expect you to script; some send a finished script to deliver. If theirs is written for the eye, ask if you can adapt the phrasing — most say yes, because sounding natural is what they hired you for.

How many hook variations should you write? Three per script is standard — brands test openers, and hooks are cheap to film since each one is its own three-second take. Write one pain hook, one curiosity hook, one result hook, same body underneath.

How long should writing a UGC script take? Twenty to forty minutes once the framework is familiar. Past an hour, you're polishing prose nobody will read — the script only has to survive being spoken.

Can you reuse the same structure for every product? Yes. The five beats fit skincare, apps, snacks, and supplements alike. What changes is the demo moment and the specifics around it — structure is the constant, proof is the variable.