We Packaged Our Proposal Workflow into a Toolkit for Other Videographers

A few months ago, we started sending clients web pages instead of PDFs. The results were immediate: a training film proposal for a professional management group closed at $10K. A recruitment film for a fire department closed at the recommended tier. A brand film for a CrossFit affiliate booked within 48 hours of the link hitting the inbox.

The proposal page isn't a fancy PDF renderer. It's a real deployed site: full-bleed hero, scroll-triggered storyboard cards, scene-by-scene shot breakdowns, tiered pricing, a production timeline. The client can leave notes on specific scenes without sending an email. Some proposals have had approval buttons that trigger when the client is ready to move forward.

We've now built enough of these that the workflow is completely repeatable. And we started getting asked about it by other videographers, so we packaged it.

What the Workflow Actually Is

The core of it is a prompt. One prompt, dropped into Claude Code, that builds the entire site: an index.html, styles.css, app.js with IntersectionObserver animations, a vercel.json, and a README. The prompt takes bracketed fields for client name, deliverables, pricing, and tone keywords, and Claude Code fills in everything else.

After the page is built, you run /impeccable, a Claude Code skill that does a full visual and code review pass. It catches inconsistencies in typography, tightens spacing, makes the color application more intentional. The difference between before and after is measurable. Pages go from functional to something that signals serious craft.

For images, we use Higgsfield's CLI. One command per image: model, aspect ratio, output path, and a cinematic prompt tied to the scene. The hero image alone communicates more about the project tone than two paragraphs of copy.

Then: git init, push to GitHub, connect to Vercel. The first deploy is live in about 90 seconds. Vercel gives you a client-name.vercel.app URL you can send immediately, or map to a custom domain in five minutes.

Why This Beats PDFs

The mechanical reason is simple: a PDF is a dead document. You cannot tell if it was opened. You cannot update it after sending. You cannot see which section the client stopped reading. A web page gives you all of that if you instrument it, and even without instrumentation, you get a live URL that you can update until the client approves.

The subtler reason is positioning. A three-tier pricing grid on a well-designed page lands differently than three rows in a table. The investment doesn't feel like a line item; it feels like a decision between options. Clients anchor to the recommended tier instead of defaulting to the cheapest.

One of the proposals in our proof set closed at a higher tier than the client initially mentioned on the phone. The page did that. Not the pitch.

What the Toolkit Includes

We published the full setup guide at /toolkit. It covers installing Claude Code, GitHub, Vercel, the /impeccable skill, and the Higgsfield CLI. The exact proposal prompt is included in full, with every section documented so you can modify it for your production style.

The prompt itself reflects real decisions we made across multiple projects: the page anatomy (hero, vision, storyboard, investment, timeline, CTA), the visual rules (#0a0a0a background, #ff6b35 accent, Inter at four weights), the reveal animation setup (IntersectionObserver with a viewport-on-load check so above-the-fold content doesn't wait for a scroll event that never fires), and the noindex, nofollow meta so proposals don't get indexed by search engines.

The Higgsfield prompt output is structured so you can run the commands immediately after Claude Code finishes, without context switching.

The Honest Tradeoff

Setup takes about 30 minutes the first time. You are installing a few CLIs and connecting a few accounts. After that, a new proposal takes 15 to 20 minutes from blank folder to live URL.

That's not zero. If you send three proposals a year, the ROI math is softer. If you send three proposals a month, or if you lose $10K jobs to videographers who look more polished, it pays for itself on the first close.

We use this for every proposal that justifies the investment. The toolkit is how we make that judgment replicable.