From First Keystroke to Team

A cinematic brand teaser begins with a lone maker typing, expands to other people and complementary skills, then resolves to a community promise and logo. The actual offer is withheld until the final third: a person starts a project; the brand helps them find collaborators.

Why it works

• A nearly black opening reveals a red keyboard and hands over the first 0.75 s; the voice lists small building blocks while the product stays hidden until 0:19.
• The cut from one typist to a wide crowd at 0:05.5, then to multiple working figures at 0:14, turns the message of collaboration into a visual progression.
• The pace compresses around 0:11–0:13 (a 0.33 s flash shot and a 1.4 s graphic), then gives the logo and promise about 6 s of quiet screen time.

Tools you'll need: a camera shoot · ElevenLabs or a recorded voice · Remotion or any video editor · Claude alone: no

Hook formula: In medias res plus a visual pattern interrupt: {{SMALL_ACTION}}, {{SECOND_INGREDIENT}}, {{THIRD_INGREDIENT}}… every {{OUTCOME}} begins somewhere.

Why it holds: The first shot is nominally 5.53 s long, but its internal motion changes from a thin light strip to hands to falling red pixels within the first 3 s. The sentence withholds where this beginning will lead until the team reveal.

Retention map: re-hooks at 0:05.5 (crowd), 0:11.1–0:12.9 (flash/graphic), and 0:15.6 (red phrase) · payoff at about 0:14–0:19 · call to action at 0:19–0:22 · hard stop on brand card, not a loop.

Mix: top-down maker close-up and digital texture; one high wide crowd shot; isolated maker scenes; red/white glitch type; floating maker collage; terminal logo card. The final 12.2 s is one measured shot with three internal changes.

Rhythm: The opening holds for 5.53 s because hands, light, and particles move inside it. Cuts quicken to 1.13 s, 0.33 s, and 1.4 s around the skill turn, then the film stops cutting and builds the end card in steps.

Rules this editor follows (inferred from the shot list):

• Start with a physical action before showing any brand.
• Expand scale from one maker to a crowd, then depict a small group with separate lit figures.
• Use one intense red/white graphic interruption before the emotional line.
• Reveal the logo in stages and leave the final lockup still enough to read.

Full shot list

• Look: Nearly black #050505 / charcoal #141414, saturated signal red about #D71920, off-white #F5F4F1, and grey #8A8A8A. Strong red practical light on hands; heavily crushed shadows; isolated figures on black. The exact source palette is estimated from the sheets.
• Captions: Closest free match Inter Regular, white, sentence case, about 0.8–1.2% of frame height (roughly 7–9 px at 720p), lower centre within the bottom 4%; little or no backing box. These are deliberately tiny and subordinate to the film. For accessible versions, raise to 24–30 px without changing placement logic.
• Graphics and overlays: Sparse red digital particles/vertical traces; briefly fragmented letterforms; one centred red sentence around 0:15.6. Closest free type matches: Inter Bold for the clean wordmark and statement; Libre Baskerville Bold for the dramatic serif/glitch card. Break or mask letters on the 0:11.5 graphic rather than copying the reference letters.
• B-roll: Top-down action in a controlled set, high-angle public crowd footage, and dark solo maker portraits; each image should convey a distinct scale or role.
• Transitions: Hard cuts dominate; the 0:11.1 flash/glitch is the main exception. Internal graphic changes punctuate the long final shot at 0:14.0, 0:15.6, and 0:19.6.
• Branding: Monogram first, wordmark second, one-line invitation third. Use {{PRODUCT}}'s own mark and type; the reference's red monogram is specific to its brand.

Measured: 127 wpm overall, 168 wpm during speech · 11 pauses of at least 0.25 s, median 0.48 s, longest 0.62 s · median pitch 131 Hz with pronounced movement (7.8 semitones standard deviation) · 6 of 8 sentence endings fall · emphasis on role changes and the later building/talent words. Opening words are unusually spaced: the first at 0:00, second about 0:01.7, third about 0:02.8.

Described: An earnest adult narrator in a medium register, measured at the opening, then more flowing through the human progression. The accent is not reliably identifiable from the supplied evidence. The pauses before new people/ideas help each picture change land.

Voice direction:

Speak as an optimistic, grounded narrator inviting makers into a shared project. Give the first three concrete nouns space, then increase pace as more people enter. Stress the role handoffs and the shift from working alone to working together. Leave a short breath before the invitation; end with a confident falling cadence. Use an original voice, not an imitation of the reference speaker.

• Music: A continuous tonal bed is measured at about −26 dBFS between words, around 6.7 dB below the voice. Its dark, low electronic character is inferred from the imagery and measurement; choose a sparse pulse or drone that can broaden subtly at the team reveal.
• Sound effects: No clear evidence of effects locked to the five cuts; the measured cut accents are weak. If adding typing, glitch, or soft impact sounds, keep them sparse and tied to visible events.
• Silence and space: Speech gaps of roughly 0.4–0.6 s separate listed ideas and the shift to the invitation; the music bed continues after the final word.
• Mix targets: Reference integrated loudness −16.7 LUFS and measured true peak +0.2 dBFS. For a new export, aim near −16 LUFS and cap true peak at −1 dBTP to avoid clipping.

• Title: A forward-looking personal promise: Your {{NEXT_CHAPTER}} starts with {{PRODUCT_OR_PEOPLE}}. The reference post uses a short invitation rather than a feature claim.
• Cover / thumbnail: Top-down pair of typing hands in a red pool of light, centre weighted against near-black surroundings; no giant cover copy in the supplied image.
• Description, hashtags, pinned comment: Post description is a single short promise plus a link; no hashtags or pinned comment are supplied.
• Audience signals: 2,524 views, 68 likes (26.9 per 1,000 views), 4 comments (1.58 per 1,000), and 4 reposts at capture. No most-replayed data or comment text is supplied, so audience reactions cannot be assigned to a specific beat.

Aim for about 43–50 spoken words across 25 s, with a spaced first list and short final lines. These are new lines for a new film, not a transcript.

[0:00–0:05.5 · SOLO SPARK · ~8 words] {{ACTION}}, {{MATERIAL}}, {{IDEA}}. Every {{PROJECT}} begins somewhere.
[0:05.5–0:11.1 · PEOPLE ENTER · ~14 words] One {{ROLE}} starts it. Another sees {{POSSIBILITY}}. Someone brings {{COMPLEMENTARY_SKILL}}.
[0:11.1–0:15.6 · TURN · ~8 words] Then the work becomes bigger than one person.
[0:15.6–0:19.1 · PAYOFF · ~9 words] You build with people who understand {{SHARED_GOAL}}.
[0:19.1–0:25.1 · BRAND/CTA · ~7 words] {{PRODUCT}}. Find {{YOUR_PEOPLE}}. Make {{NEXT_OUTCOME}}.

Toolkit

• Live footage: a camera shoot. Films makers, hands, and the growing group
• Voice: ElevenLabs or a recorded voice. Narrates the path from maker to team
• Edit: Remotion or any video editor. Shapes the montage and staged brand reveal

Mostly filmed: Claude can plan it, design the graphics and cut it, but the footage needs a shoot. Red lighting and casting carry the emotional shift.

Script prompt (paste into any AI assistant with this file attached):

Write an original 25-second brand teaser for {{PRODUCT}}, serving {{AUDIENCE}} around {{TOPIC}}. Use about 43–50 spoken words. Begin with three spaced, tangible building blocks and a maker doing a physical action. At 0:05.5 widen to other people; at 0:11 introduce a complementary skill; at 0:13–0:19 reveal the emotional value of working together. Save the product name and one action invitation for 0:19 onward. Use short falling sentences, no questions, no feature list, and no phrases from the reference. Return a timed script, on-screen text, and a distinct visual for every role.

Voice prompt:

An original adult narrator, medium register, grounded and optimistic. Start with deliberately separated words, then speak around 165–170 wpm during the main body. Short pauses mark each new role and a longer breath precedes the team insight. Clear final invitation; no celebrity or creator imitation.

Visual prompts:

• Hands / action: Top-down close-up of anonymous hands performing {{ACTION}} on {{TOOL}}, practical saturated red light, almost black room, textured shadows, cinematic but natural movement, 16:9.
• Crowd / scale: High wide view of varied anonymous adults crossing an open public space, restrained black-and-white grade, clear tiny silhouettes, 16:9.
• Makers / collage: Three to five anonymous makers working on different parts of {{PROJECT}}, isolated pools of pale and red light against black; separate windows and negative space for centred copy, 16:9.

Caption spec: Inter Regular, tiny white sentence subtitles, bottom centre; large statement in Inter Bold red at the turn. In an accessible export, substitute 24–30 px subtitles.

Music prompt: Sparse, dark electronic ambient bed with a restrained low pulse, no melody competing with speech, slight widening near 0:14, resolves without a hard sting. Seek a licensed or original cue.

1. Lock a 25 s timeline at 1280×720 and 15 fps (or preserve these time ratios at 24/30 fps).
2. Record original voice at roughly 46 words; place deliberate gaps after the opening list and before the team/brand lines.
3. Build the top-down red action shot first; animate its reveal and red particle transition without adding an early brand title.
4. Cut to wide people at 0:05.5, solo maker(s) at 0:06.7, a 0.33 s flash at 0:11.1, and a 1.4 s red skill card at 0:11.5.
5. Start the black-background maker collage at 0:12.9. Add its red insight at 0:15.6, brand symbol at 0:19.6, then wordmark and invitation by 0:22.
6. Add tiny bottom subtitles, the continuous bed, and only motivated accents. Grade all footage toward deep black, signal red, off-white, and muted grey.
7. Export and check logo read time, subtitle readability, and peak level.

• The first human action is visible by 0:00.75; the opening uses motion even though the first cut waits until 0:05.5.
• The visual scale moves from one person to many by 0:06.
• A red/white interruption lands around 0:11–0:13, followed by an emotional team statement near 0:15–0:19.
• Target 6 shots, about 5 cuts, and 3 internal reveals; final lockup is readable for at least 2 s.
• Voice is about 165–170 wpm while speaking, with a spaced opening and clear role handoffs.
• Small subtitles are reviewed for readability at phone size; accessible version uses larger type.
• Audio is near −16 LUFS integrated and no higher than −1 dBTP true peak.
• All footage, logo, copy, and voice are original or properly licensed.

• The creator's signature (don't copy): YNS's exact logo/monogram, red typographic mark, script, and linked invitation.
• The transferable format (do copy): A lone physical action, expansion to multiple roles, a fast mid-film skill flash, a team insight, and a staged brand reveal on a limited palette.
• Replace the roles, action, promise, and ending with the actual world of {{PRODUCT}}; use your own footage, voice, music, and brand assets.

Source: https://x.com/ynsgh_/status/2113670281600418046 · measured with dissect 1.1.0 on 2026-10-26 · descriptions marked “inferred” are judgement calls.