Zoom out to the big picture or zoom in on the detail to find a better way to frame the problem.
Any problem can be seen at different sizes. “Our checkout is slow” is part of something bigger (people getting what they came for) and is made of something smaller (three screens, a payment step, one form that's painful on a phone). Chunking means moving between these levels on purpose instead of staying stuck at the one where the problem first appeared.
Chunk up by asking “Why does this matter?” or “What is this an example of?” to reach the bigger picture. Chunk down by asking “What, specifically?” or “Which part?” to reach concrete detail. Then try the creative move, chunking sideways: climb to the general category, and come back down to a different example of the same thing.
Use it when a problem feels too big to start on, when it's framed so narrowly that only one answer fits, when a team is lost in detail, or when you need a fresh angle on a familiar issue.
Start in the middle. Zoom out to see what it's part of, zoom in to see what it's made of.
Choose a part of the diagram or a tab to explore each level.
What is this ultimately part of, or for?
What is this an example of?
How would you describe the problem right now, in one line?
What, specifically, is it made of, and which part matters most?
Can you give me a specific example?
People tend to get stuck at one level of detail. Chunking moves you on purpose.
Some people dive straight into detail; others live among big ideas. A team stuck at one level keeps producing the same kind of answer. Moving up and down gives you access to both kinds of thinking.
A problem like “improve student well-being” is too big to act on. Chunking down finds a piece small enough to start on this week, without losing sight of why it matters.
Climb to the general category, then come down to a different example, and you've found an analogy. If your problem is “friction before a purchase”, how do restaurants, airports and libraries deal with the same kind of friction?
A problem stated too narrowly allows only one answer. Chunking up reveals the purpose behind it, which opens up options the narrow version had quietly ruled out.
Write the problem in the middle of a page, in plain words. This is your starting level. Leave room above and below it.
Ask “Why does this matter?” and “What is this an example of?” Write each answer on the level above. Go up two or three levels, until the problem feels broad but still meaningful.
Go back to the start and ask “What, specifically?” and “Which part?” Write the answers below. Keep going until you reach something concrete you could observe, count or try this week.
From a higher level, ask “What else is an example of this?” List three or four parallel cases, ideally from other fields. Each one is a source of borrowed ideas.
Look at the whole ladder and pick the level where you can act and still have room for many ideas. Write it as a new Problem Statement. The best frame is often a level or two away from where you started.
A page with a ladder sketched on it, one problem and about 20 minutes. It works well with a partner who asks the questions. With AI, use the prompt below.
Subject: a family of four deciding whether to replace a 14-year-old car. Goal: reliable ways to get around without wrecking the budget. Constraint: about $9,000 saved.
A family life where everyone gets where they need to go, safely, without constant money worries.
Getting four people to work, school and practice every week.
“Which new car should we buy?”
One 12-mile commute, two school runs a day, weekend sports and one long road trip a year.
Of about 30 trips a week, most are under three miles. The commute has a bus with one transfer.
How else do families solve “getting people around”? Carpools, e-bikes, car sharing, renting for the long trips.
The question changed from “Which car should we buy?” to “How do we cover 30 short trips a week and one long one?” The family kept the old car one more year, bought an e-bike for school runs and joined a school carpool, keeping most of the savings.
Chunking is easy to understand and surprisingly hard to do alone, because we keep sliding back to our favorite level. An AI partner can keep asking “why does that matter?” and “what, specifically?”, and it can suggest parallel examples from other fields when you go sideways. You decide which level is real and useful.
Build the whole ladder, one question at a time.
Ask me one question at a time to chunk my problem up two levels and down two levels.
Find analogies by changing the example, not the category.
My problem is an example of [category]. Give me 6 other examples from different fields, and what each teaches.
See the problem restated at every level of the ladder.
Rewrite my problem as a "How might we...?" question at each level of this ladder: [paste ladder]
Get a second opinion on which level to work at.
Which level of my ladder best balances room for ideas with the ability to act? Explain why.
Describe the problem as it looks to you right now, then copy the prompt into any AI assistant. In coach mode the AI moves you up, down and sideways one level at a time, waits for your answers and helps you choose the frame worth working on.
The “Open in” links put the prompt in the web address. If your challenge includes anything sensitive, copy and paste instead.
After the session, click one to copy it, then paste it into the same chat.