October 23, 2026

Studio: Landing Pages at Cents per Venture

Every venture experiment starts with a page: something a potential customer can land on, read, and sign up from. Until now our ventures were built with Lovable, an excellent product builder but an expensive way to make a page that may only live for two weeks.

This week Studio, our own page builder, became the default for venture pages. Lovable stays in the picture, but only for the full product, and only once an experiment has shown there is demand.

1. What Studio Is

Studio gives every venture its own real codebase and an agent that works on it the way a developer would:

  • One Project per Venture — A real React and TypeScript project, with Tailwind and shadcn/ui, in its own git repository.

  • Work in Turns — Each request, such as "build the page" or "make the hero darker", is one turn. The agent edits the code, the code is type-checked, and the page is rendered.

  • Every Turn Is a Version — Each turn is saved as a version that can be previewed, published, or restored later.

  • One Fast Static Page — The result is a single static HTML page with nothing to load on top, served at the venture's own address.

2. Landing Pages First

Each page carries an email signup and the ad tracking a campaign needs. An experiment runs a small ad budget against it with a clear goal and a time limit. Only an experiment that meets its goal unlocks the full product build. Building the product is the expensive part, so we now do it only for the ventures that earn it.

3. Every Venture Its Own Look

The first Studio pages all looked alike. Each venture now gets an art direction from a set of twelve, picked to fit the business, and directions given out recently are not offered again. Every page also gets three or four generated pictures, for about thirteen cents per venture.

4. Cents, Not Dollars

In our bake-off earlier this month, our open-source design pipeline produced pages good enough to ship for two to four cents each, in under five minutes. Our past Lovable builds took about fifteen minutes and cost more per build before counting the plan's credits. Studio runs on the same low-cost models, and every turn has a hard spending cap.

Weekly Learnings

  • Validate before you build — A page and a small ad budget tell us whether anyone wants a product, long before it is worth building one.

  • A page is a test, not a product — Landing pages need to be good, fast and cheap, and they can be thrown away. Products need care. They deserve different tools.

  • Variety matters — Pages built from one template look like one template. Art directions and pictures made the difference.

Plan for Next Week

  • Close the Quality Gap — Bring Studio's pages up to the standard of Lovable's, combining it with the open-source design pipeline.

  • From Plan to Product — Feed First Pancake's full product spec into Studio to build working products, not only pages.

  • Pause the Expensive Plan — Hold the Lovable subscription until a venture validates and needs a full build.

Subscribe via email to follow Entonomy updates: