October 21, 2026
Batch Mode: Ten Ventures From One Click

This week we turned starting a venture into a batch job. Until now every idea went through First Pancake on its own: one plan, one brand, one landing page, and then the next. Batch mode takes up to ten ideas at once and builds all of them side by side.
The aim is simple. If AI agents can run a business experiment end to end, the number of experiments we can run should not be limited by how many times a person clicks Start.
1. Ten Ideas, One Start
A batch starts from a single page. You type your ideas or tick them from the Idea Hub, choose how much effort the agents should spend, and press Start. Signed-in users can start ten ideas at a time; visitors without an account can try six. Every idea in the batch gets the full package:
A 24-Step Business Plan — Every step of MIT's Disciplined Entrepreneurship method, from market segmentation to the product plan.
A Brand — Name, colors and type for the venture, made before anything is published.
A Landing Page — A real page with an email signup, ready for its first ads.
Eight More Tactics — The rest of the go-to-market kit, for ten tactics in all.
2. Built in Parallel, Not in a Queue
Our first version of batch mode was quietly sequential: the steps of the first plan took every available worker, so idea two waited for idea one. Each plan now gets its own share of the workers, so all ten move forward together. Every idea has a card showing its stage, the steps being written, and what it has cost so far.
3. One View of Every Venture
A batch creates many ventures at once, so they now share one overview: newest first, with a Just created card for the last batch. The same page has a Stop all switch. One click pauses every venture, stops strategy runs, and turns off live ad campaigns and email sequences, so nothing keeps spending while we look at the results.
4. Ideas Straight From the Idea Hub
The Idea Hub scans trend sources every day (search trends, launch sites, communities) for problems people are actively reporting, and scores each one as a business. The batch page puts those ideas one tick away, so a batch can start from what the market is asking for rather than what we happen to think of.
Weekly Learnings
Parallel beats sequential — Starting ten ventures should take the same effort as starting one. The bottleneck is now how good the ideas are, not how fast we can start them.
A venture is only as good as its idea — Ten experiments on weak ideas teach us less than three on strong ones. Idea quality is now the input that matters most.
Every fleet needs an off switch — Running many ventures at once only works if we can stop all of them at once.
Plan for Next Week
Bigger Batches — Raise the batch limit from ten to twelve ideas.
Better Ideas — Add more sources to the Idea Hub and score them all with one rubric, framed around real customer pain that could support a long-lived business.
A Traction Score — Show on the overview which ventures are getting real responses, not only what each one has spent.