LaunchCut

I’ve been building LaunchCut, a tool that takes a landing page URL and generates a short launch video for places like Product Hunt, X, Reddit, or a product homepage.

The basic flow is:

1. Paste in a landing page URL

2. Extract the core messaging, screenshots, visuals, and product context

3. Generate a script/storyboard

4. Turn that into an editable video

5. Let the founder polish and export it

A few things I’ve learned so far:

1. Landing pages are surprisingly inconsistent

Some pages explain the product clearly in one hero section. Others bury the actual value prop halfway down the page, or rely heavily on screenshots, carousels, animations, or vague copy.

The hardest part has been understanding what the product actually does from the landing page, then deciding what deserves attention in a short video.

2. “Generate a video” is too broad

The more useful framing seems to be: generate a solid first draft that a founder can quickly polish.

People still want control over messaging, pacing, visuals, and tone. Full automation sounds nice, but for something public-facing, editability matters a lot.

3. The storyboard matters more than the video rendering

The biggest quality jump has come from improving the structure of the video: hook, problem, product, benefits, proof, CTA.

If the story is weak, better visuals do not save it.

4. Product videos are not just launch assets

I originally thought this would mostly be useful for Product Hunt launches, but I’m starting to think the bigger use case might be ongoing content: feature announcements, demo clips, changelog videos, landing page explainers, and social posts.

Tech stack

The app is built with Next.js/React. Video generation is handled with Remotion.

For the AI layer, I’m using LLMs to analyze the landing page and generate the video script/storyboard. The tricky part is making the output structured enough that it can reliably map into video scenes, rather than getting a generic marketing script.

Curious how other builders would think about this:

1. Would this be more useful for one-off launches, or for ongoing content?

2. If you were building this, where would you put the most effort: scraping, script generation, editing UX, templates, or rendering quality?