Skip to main content
This guide walks through building a real planning tree, from a vague idea to a structured strategy with delivery-ready jobs.

​
Start with what you know

Let’s say you’re building a recipe sharing app. You don’t need a business plan. Start with what’s on your mind:
“I want people to share recipes and discover new ones”
Type this into the +Need chat. The AI will help you figure out where it fits.

​
The AI maps your thinking

The AI takes your input and structures it:
  • Goal: Build a recipe platform that grows through sharing
  • Need: Users can discover recipes that match their preferences
  • Need: Users can share their own recipes easily
Review what it suggests. Edit labels, remove what doesn’t fit, add what’s missing. Then commit.

​
Expand with approaches

Click on a need and add approaches. Specific bets on how to address it. For “Users can discover recipes that match their preferences”:
  • Approach: Tag-based filtering (must-have, skateboard)
  • Approach: AI-powered taste matching (delighter, car)
Each approach has a measure. How you’ll know it works:
  • “Users find a recipe they save within 3 searches”
  • “AI recommendations get 40%+ save rate”

​
Generate jobs

Select an approach and click Propose jobs. The AI generates acceptance criteria: For “Tag-based filtering”:
  • “Filter panel shows available tags from recipe database”
  • “Selecting a tag narrows results to matching recipes”
  • “Multiple tags combine with AND logic”
  • “Clear filters resets to full list”
Each job describes what you’d see if it’s working. Not how to build it.

​
Create an iteration and assign

Create an iteration called “MVP”. Assign the must-have jobs. Switch to the Delivery Map to see your first release plan, approaches as columns, your MVP as a row, with the proof you need for each bet. You now have a structured strategy that traces from business value through user needs to specific, verifiable delivery checkpoints.