# Start from a Template: first-fold video script 40 second hero cut. Screen recording, no face, no hands. Built for muted autoplay and loop, so the on-screen text carries it and the VO is only for the long cut. Grounded in the JAS '26 handoff: `figma.com/design/OywpgPNGvoFAIS4haGGehy/JAS--26---Handoff?node-id=1-2798` ## The claim A template is a position on what an excellent version of this deck answers, handed to you before you start. The job to be done is not "give me an outline", it is *make sure this is good and I haven't missed anything.* The fear is walking into the room having skipped something and finding out when someone asks. Two rules fall out of that, and they govern every line of copy: - **Never say "we".** The template might be one the user's own team saved to govern their proposals. Write the authority as a property of the template, not of Presentations.ai. - **Never show a slide count.** Counting slides drags the claim back down to mechanism. Coverage is the point, quantity is not. ## Cast | | | |---|---| | Persona | Anika Bose, Marketing Lead at **Marlow** (fictional cold brew brand) | | Pitching | **Trailhead**, a fictional outdoor gear brand | | Template | Brand Partnership Proposal, Marketing tab | | Her campaign idea | "The Early Start" | ## Storyboard ``` ┌─────────┐ ┌────────┐ ┌──────────┐ ┌─────────┐ ┌────────┐ ┌────────┐ ┌────────┐ │dashboard│→ │gallery │→ │ FILMSTRIP│→ │braindump│→ │shimmer │→ │ EDITOR │→ │ end │ │ use a │ │Marketing│ │ what it │ │ pasted │ │ + trace│ │nothing │ │ card │ │template │ │ tab │ │ answers │ │ │ │ │ │dropped │ │ │ └─────────┘ └────────┘ └──────────┘ └─────────┘ └────────┘ └────────┘ └────────┘ 0:00 0:03 0:07 0:14 0:22 0:26 0:36 ── what good looks like ── ──── her context, built on top ──── ``` ## Shot by shot | Time | On screen | Text on screen | VO (long cut) | |---|---|---|---| | **0:00 to 0:03** | **Dashboard.** Cursor skips the prompt box and clicks **Use a Template** in the quick actions row. | | "You don't always want to start from a blank prompt." | | **0:03 to 0:07** | **Gallery.** Select a template, full screen. Cursor taps the **Marketing** tab, the grid reflows, lands on `Brand Partnership Proposal`. Hold on the tabs about a second: they are what says this is a library of deck types, not colour schemes. | | | | **0:07 to 0:14** | **The authority shot.** Detail lightbox opens. Camera moves down to the filmstrip and scrubs it, slide titles readable as they pass: *Where our audiences overlap · The value exchange · How we will measure it · Next steps.* Hold on the last one. | `Someone already worked out what a good one answers.` | "Someone has already worked out what a partnership proposal has to answer to actually land." | | **0:14 to 0:22** | **The braindump.** `Use this Template` opens the split screen. Right: the template's slides, fully designed and populated. Left: the headline `Share your own context, and we'll update the template`, and her notes pasting in, scrolling as they grow. Visibly a mess. `Create Presentation` turns navy. | `You bring what you know.` | "You bring your side of it, in whatever state it's in." | | **0:22 to 0:26** | **Simmering.** A shimmer sweeps across the slides on the right, a soft highlight moving over the blur rather than a static frost, about one pass every 1.2s. Left shows the trace: *Developing Presentation Outline · Refining Strategic Narrative · Analyzing Partnership Value.* `Letting it simmer… 33s`. | | | | **0:26 to 0:31** | **The editor.** Cut. Slide rail on the left, the full deck, same order as the filmstrip. Camera holds on the rail a full beat before touching the canvas, so the viewer registers that nothing was dropped. | `Nothing left out.` | | | **0:31 to 0:36** | **The question she skipped.** Canvas scrolls to **How we will measure it**, filled: bundle units, joint email CTR, combined reach, new-to-file customers each side, cost per new customer. There is not one word about measurement anywhere in her notes. | `Including the question you'd have been asked.` | "Including the ones you'd only discover you'd skipped when someone asks." | | **0:36 to 0:40** | **End.** Logo, one navy CTA. | *Don't start from blank.* **Start from what good looks like.** | | **0:07 is the shot that cannot be cut.** It is the only moment in the product where you see what the deck sets out to cover before any of the user's content exists. If the titles are not readable at hero scale, the video is just another generation demo. ## What she types The most important piece of writing in the video. Lowercase, uneven, open questions left in it. If it reads as a well-formed prompt, the contrast dies. ``` ok trailhead partnership thoughts before i forget their spring gear drop is apr 14, our cold brew season basically starts the same week, feels too obvious not to do audience, we're 1.4M monthly, they're 2.1M, overlap is maybe 380k? ask growth for the real number they have 40 stores and a 900k list which is the thing we actually want. we have capacity (12k units) and the studio idea: "the early start". 5am, first light, first mile, first cup. bundle is their flask plus our beans no cash either way, in kind, media value probably evens out around 180k each side but finance needs to confirm ridge cycles worked, 8.4k bundles in six weeks, 22% new to us, use that legal will want 3 weeks so kickoff has to be feb who owns this on their side? need a name before the meeting ``` ## What the template makes her answer Full content in `brand-partnership-proposal-template.md`. Bold rows are the four her braindump never touches, which is where the template earns its place. | Slide | The question it forces | |---|---| | The opportunity | Why is this worth doing, and why this quarter | | Where our audiences overlap | Do our audiences actually meet, and where | | What we bring | What are we putting in | | What you bring | What are we asking of them, said plainly | | The campaign | What is the actual creative | | **What we would make together** | What gets made, and where does it run | | **Timeline** | When does each side have to move | | **How we will measure it** | How will we both know if it worked | | The value exchange | Who puts in what, and is it even | | We have done this before | Why should they believe we can run this | | **Next steps** | What happens after the meeting, and who owns it | ## Production notes - **Match the filmstrip order to the editor rail exactly.** One slide moving between those two shots collapses the whole argument. This is the only continuity error that matters. - **Keep the shimmer beat.** Cutting to the editor early reads as a scene change rather than the same deck arriving. - **Specific, odd numbers** (`8.4K bundles`, `roughly 380K`). Round numbers read as placeholder. - No exclamation marks. One navy CTA at the end. Swap Marlow and Trailhead for whatever clears legal. ## Open dependencies **The filmstrip is currently unreadable at hero scale**, small and cropped, so the shot the video depends on does not exist yet. That panel probably wants a slide-outline list beside the thumbnails. Worth doing regardless of the video: the outline is the reason to pick one template over another, and right now you cannot see it before committing. **The template has to be built.** Its copy is written: card, description and all twelve slides are in `brand-partnership-proposal-template.md` beside this file. It does not exist in the Marketing tab yet. **Unresolved: does a template ship filled or prompted?** The 0:14 shot above describes the right panel as fully populated, which is what the handoff mock shows. The template doc specifies headings and prompts instead, with the filled version existing only as the gallery thumbnail. Both are defensible and they look completely different on screen for the middle twenty seconds of this video. Settle it before either gets built. ## Not in this video The team-authored template case, where a team saves their own so every proposal they ship covers the points they have decided matter. Stronger enterprise story, but it needs a second persona and a governance beat that 40 seconds cannot hold. Its own video.