PROD. SECOND UNIT UGC/SIDES — MODULE 2/DRAFT/Fact-check before publish

demo-review30–60s typicalModule 2  ·  section 9.2  ·  1,116 words · ~6 min read

Template — Demo / review

templateDoc type
draftStatus
RequiredFact-check
neverLast verified

1What it is

The product doing the thing it's for, on camera, with you narrating. The centre of gravity is the object, not your face. A testimonial asks the audience to believe you; a demo asks them to watch and decide for themselves.

Typical length: 30–60 seconds. The longest of the four formats, because it has to actually show something happening in real time.

2When it's the wrong format

  • When the product's benefit isn't visible. Software, supplements, anything where the result happens elsewhere or over weeks. Forcing a demo produces footage of someone holding a bottle.
  • When you can't demonstrate it correctly. Showing incorrect use is a problem in every category and a serious one in some. If you're not sure of the right method, ask the brand before you shoot, not after.
  • When the brand wants a pure endorsement. They briefed a testimonial. Give them a testimonial.

3The beats

Beat sheeta beat is a job, not a number of seconds
1The claim under testState what it's supposed to do, in the brand's own words~10%
2SetupConditions. What we're using, on what, in what state.~15%
3The doingReal time, unbroken if possible. The reason the piece exists.~45%
4The resultHeld long enough to be read.~20%
5Verdict with a limitDid it do the thing, and where doesn't it.~10%

The doing beat should resist cutting. Every cut inside a demonstration invites the thought that something was skipped, and that thought is fatal to the format. If a process genuinely takes four minutes, show a continuous chunk and timestamp the jump honestly rather than dissolving through it.

The setup beat is what separates a demo from an advert. Naming the conditions — cold pan, no oil, straight from the fridge — is what makes the result mean anything. Skipping it is the single most common reason a demo reads as staged.

4The brief

Brief  demo-review30–60s typical
Write a [45]-second demo script in [5] beats.

PRODUCT
  [PRODUCT NAME][ONE PLAIN SENTENCE: what it is]
  Manufacturer's claim being tested: [THE SPECIFIC CLAIM, IN THEIR WORDS]

CONDITIONS — these appear on screen, so be exact
  [WHAT I'M TESTING IT ON]
  [STARTING STATE — temperature, condition, quantity, whatever is relevant]
  [ANYTHING THAT WOULD MAKE THIS UNFAIR IF OMITTED]

WHAT ACTUALLY HAPPENED — do not improve on this
  Result: [WHAT I OBSERVED, INCLUDING TIMINGS IF RELEVANT]
  Where it fell short: [THE LIMIT, or "none observed"]

AUDIENCE
  [ONE PERSON IN A SITUATION]

BRAND VOICE
  Never say: [OFF-LIMITS LIST FROM FORM BV-01]
  Barred claims: [BARRED CLAIMS FROM FORM BV-01]

CONSTRAINTS
  - Beat 3 is continuous — write it as one unbroken action, no cuts
  - Narration under 12 words per beat; the picture is doing the work
  - No comparative claims against named competitors
  - State the conditions out loud in beat 2, do not leave them to captions alone
  - Do not claim a result I did not list under WHAT ACTUALLY HAPPENED

Return as: BEAT n — [narration] / [what's on screen]. Nothing else.
Everything in [SQUARE BRACKETS] is yours to replace. Everything else stays.

The narration word limit is the important constraint. Beginners over-narrate demos, and the format's whole advantage is that the viewer can see it. If the words and the picture say the same thing, cut the words.

No comparative claims against named competitors is in there because it's a common contractual bar and an easy accidental breach — "much better than the one I had" is fine; naming the one you had usually is not. Check Form BV-01.

5What to reject

  • Any result not in your WHAT ACTUALLY HAPPENED block. Same discipline as the testimonial, higher stakes, because a demo implies evidence.
  • Cuts inside the doing beat that the script sneaks back in.
  • Narration that describes what's visible. "So I'm pouring it in now" — the audience can see you pouring it in.
  • A verdict with no limit. In this format especially, a flawless result reported flawlessly reads as an advert, and viewers discount it accordingly.
  • Conditions that quietly favour the product. If you preheated the pan and don't say so, the demo is dishonest even though every individual frame is real.

6Illustration

REAL GENERATED OUTPUT

Produced 2026-08-14 by claude-opus-5 via ../skills/script-from-intake, from the demo intake in ../skills/demo-runs/_inputs.md. The full run — brief, first pass, claim-guard report and every human edit — is in ../skills/demo-runs/02-demo-review.md.

The brand and the creator are invented and clearly labelled as such. The output is genuinely generated; the inputs are not a real client. That makes this a real example, not a case study.

Generateddemo brand
Output  real generated outputdemo brand — invented, labelled
DISCLOSURE — "Paid partnership with Sable."          / title card + spoken, 0:00–0:02
BEAT 1 — "Sable says eggs come off with almost no oil." / pan held, one beat
BEAT 2 — "Cold pan. Smear of oil. Fridge-cold eggs. Pan's four weeks old."
                                                        / hand places each item
BEAT 3 — [no narration — run continuous, do not cut]    / egg in, cooking, spatula under
BEAT 4 — "Came off in one piece."                       / empty pan, held three seconds
BEAT 5 — "Heavy. Hand wash only. It did the thing."     / pan set down, hand out of frame

Beat 3 has no narration and beat 4 holds on an empty pan — the absence of residue is the result, and it needs time to register.

The first pass of this script failed a check, and it is worth knowing how. Beat 3 originally read "about ninety seconds of cooking". The creator had never timed it; the experience file said so explicitly. The generator's own self-check passed the line because it scans narration, and the timing sat inside a stage direction. claim-guard caught it on the claim-to-experience trace. It was removed rather than softened to "a minute or so" — a hedged unsupported claim is still unsupported, and it is harder to catch on the second read.

7Adapting this

  • Multi-step products: split beat 3 into two continuous chunks with an honest on-screen time jump between them, rather than four short cuts.
  • Comparative demos (this vs. what you had before): run beats 2–4 twice, identical conditions both times, and say out loud that the conditions were identical. Confirm the brand permits comparison first.
  • Regulated categories — anything health, financial or safety-related: do not use this template without the brand's own approved claim language in beat 1, and expect a compliance review. This is not a place to improvise.
Your word for it — nothing is tracked automatically.

Source: content/module-2/templates/02-demo-review.md