Ask “what is it for?” again and again until you reach the purpose behind the request.
Problems often arrive dressed as requests: “We need a new website.” “Can we get a bigger room?” Purposing treats the request as the bottom rung of a ladder and climbs it with one question: “What is it for?” Each answer reveals a purpose, and then you ask the same question of that purpose.
A few rungs up, you usually reach something more important than the original request, and more interesting to solve. From there you can come back down by a different route. Once you know the new website is really meant to help families sign up faster, a shorter paper form at the front desk might matter more than any redesign.
Ask from more than one viewpoint, too. The person who asked, the people who'll use the result and the people who'll pay for it may each have a different answer to “what for?”
Use it when someone hands you a solution as if it were the problem, when a project has lost sight of why it started, or when a team is stuck arguing about how before agreeing on why.
Climb with “What is it for?” Then come back down by a different route: other ways to serve the same purpose.
Choose a part of the diagram or a tab to explore each rung.
And why does that matter, in the end?
What would be different if the need were met?
What need is this meant to meet?
What will it be used for, exactly?
What exactly was asked for, in the asker's own words?
It's one of the simplest tools on this site. It works because of the way requests usually reach us.
People rarely bring you a need. They bring the first answer that occurred to them. Climbing to the purpose separates the need from that first answer, so you can look for better ones.
Most purposes can be reached in many ways. Once you're a rung or two up, you can see alternatives that the original request ruled out without anyone noticing.
Two people can ask for the same thing for different reasons, or for different things for the same reason. Talking about purpose finds common ground faster than arguing about features.
Too low, and you're just filling an order. Too high, and the problem becomes vague. Seeing all the rungs at once lets you choose the level where you can make a real difference.
Write the request at the bottom of a page, in the words of the person who asked. Don't improve it yet. The exact wording often reveals the solution that has already been assumed.
Write the answer on the rung above, then ask the question again about that answer. Keep climbing three to five times, until the answers become broad values that everyone would agree with.
Who else cares about this, and what is it for, from their side? The user, the payer and the person who will maintain it may give different answers. Add them as side branches.
Pick the level that is high enough to open new options and low enough to act on. That is your real problem. Write it as a goal or as a How To statement.
From your chosen rung, ask “How else could we achieve this?” and list at least five routes, including the original request. Then compare them honestly.
A page with a ladder sketched on it, the original request and about 15 minutes, ideally with the person who asked. With AI, use the prompt below.
Subject: a middle school's request for more classroom laptops. Goal: decide how to spend a $12,000 technology budget. Constraint: the money must be spent this school year.
“We need 60 more laptops.”
So every student in a class can be online at the same time.
So teachers can run reading and math practice software during class.
So each student practices at their own level, and teachers can see who is struggling.
So no student falls behind without someone noticing in time.
Rung 4: practice at each student's level, with clear signs of who is struggling. Open enough for new options, concrete enough to act on.
From rung 4 the team mapped new routes: buy 20 laptops and run practice in station rotations, so only a third of a class needs a device at once; bring 30 older library tablets back into use; and fund a part-time tutor with what's left.
Purposing needs someone who keeps asking “and what is that for?” long after it feels polite. AI can do that without awkwardness, and it can suggest purposes you hadn't considered and routes back down you'd never think of. You know which purposes are real, and which rung is worth working on.
Have AI ask the question one rung at a time, building on your answers.
Ask me "What is it for?" about my request, one rung at a time, and redraw the ladder as we climb.
See how other people might answer the same question.
How would a user, a funder and the person maintaining it each answer "What is it for?" about my request?
Generate alternatives that serve the purpose you chose.
My real purpose is [chosen rung]. Give me 10 ways to achieve it, including some that skip the original request.
Test whether the rung you chose is the right level.
Is my chosen purpose too broad to act on, or too narrow to allow options? Suggest a better wording.
Write the request exactly as it was made, then copy the prompt into any AI assistant. In coach mode the AI asks “what is it for?” one rung at a time, waits for your answers, then helps you pick the right rung and find new routes back down.
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.