Viral Content Script Generator + Framework

The constraints

What made this hard?

Three constraints. The first was retention structure. Short-form video on Instagram, LinkedIn, and short-form video platforms generally rewards a specific shape: hook in the first three seconds (otherwise viewers swipe), body that delivers on the hook, a mid-video retention line that gives viewers a reason to keep watching, and a conclusion with a payoff or a call. Most founder-shot video misses two of those because the structure isn't named.

The second was founder-as-creator. Most short-form advice assumes the creator wants to be a creator. Founders don't; they want their company to be on the platform. The framework had to assume the creator is reluctant, busy, and not a natural performer, and produce scripts that work for someone reading off a phone teleprompter rather than improvising.

The third was the distribution layer. A great script that gets posted into a void doesn't work. The framework had to cover not just scripting but distribution and repurposing, so the founder isn't relying on platform algorithms alone for reach.

The approach

How did Tincture frame the problem?

Two pieces, designed as a system. The Framework documents the seven-step methodology: plan (pick the topic and the hook angle), script (structure the hook, body, mid-video retention, conclusion), film (the filming setup, tone, and approach for non-creators), edit (cuts, captions, retention edits), publish (platform-native versions per channel), distribute (cross-posting, repurposing into LinkedIn and other written surfaces), measure (which retention markers to watch, which formats to repeat). The framework names each step with the specific thing the founder has to decide rather than the philosophical principle.

The Generator runs the scripting step of the framework as a Claude-powered tool. Topic in, complete short-form script out: hook (with three or four variants to pick from), body (structured beats), mid-video retention line, conclusion, plus filming notes (how long, what to cut to, what to emphasise). The output is teleprompter-ready, formatted so a founder can read it off a phone and shoot.

The pairing matters. Founders who use only the Generator without the Framework get scripts that don't fit their distribution motion. Founders who use only the Framework without the Generator stall at the scripting step every time. Both shipped together because that's where they earn their keep.

The build

What was shipped?

The Framework: a documented seven-step methodology covering plan, script, film, edit, publish, distribute, measure. Each step has the specific decisions to make, the rules of thumb (hook in three seconds, retention line at the midpoint, payoff before the call), and worked examples. Distributed as a free guide with a clear structure.

The Generator: a web tool delivered as an overlay, with topic and angle as inputs, a Claude API call running the scripting layer of the framework, and the output rendering as a complete short-form script. Hook variants to pick from, structured body, mid-video retention line, conclusion, filming notes. Copy-able to clipboard, downloadable as text.

The Generator's prompt scaffolding encodes the framework's scripting principles directly. The tool produces framework-quality output without requiring the founder to internalise the framework first; the framework deepens the founder's understanding of why the script is shaped the way it is.

The outcome

What were the results?

Both shipped, both free, both live. Founders use the Framework to set up their distribution motion and the Generator to script individual videos. The combined effect: founders who were stalled on "I should be on short-form" ship their first ten videos in the next month rather than the next quarter.

The compounding outcome is the structural learning. After scripting half a dozen videos using the Generator, the founder internalises the hook-body-retention-conclusion shape and starts writing better hooks unprompted. The tool teaches the framework by running it; the framework deepens the tool's outputs.

The strategic outcome is the retention shift. Founder-shot video that follows the structure retains viewers materially longer than founder-shot video that doesn't. That's the whole point. Distribution surfaces reward retention; retention rewards structure; structure is the script.

What it took

What tools and methods were used?

The Framework: documented as a free guide with worked examples and the seven-step structure. The Generator: web app delivered as an overlay, Claude API call running the scripting layer, structured output formatted for teleprompter use.

The methodological underpinning is the practice's pattern for paired methodology-plus-tool products: ship both because the methodology without the tool stalls at execution, and the tool without the methodology produces output the founder can't extend. The pairing teaches and ships at the same time.

The other move worth naming: structure as the script. The hook-body-retention-conclusion shape isn't an abstract idea about good content; it's the mechanical reason short-form video retains attention. Encoding that structure into the prompt scaffolding rather than hoping the founder gets it right at scripting time is what makes the tool's output reliable.

The takeaway

What's the transferable principle?

Most short-form video content fails at retention, not at the idea. The hook isn't strong enough, the body wanders, the mid-video retention line isn't there, the conclusion fizzles. The work that lands names the structure that holds attention and runs the scripting against it.

For the Framework and Generator, that meant pairing the methodology (seven steps, named decisions) with the execution tool (Claude-powered scripting that follows the structure). Founders who use both ship videos that retain better than founders who pick up the camera with vibes.

The other transferable principle, broader than short-form: paired methodology-plus-tool products land harder than either piece alone. The methodology without the tool stalls at execution; the tool without the methodology produces output the user can't extend. Ship both, design them as a system, and the system compounds in ways either piece individually can't.

Read more on this in LinkedIn: A Distribution Strategy That Doesn't Require Becoming a Thought Leader