Data & Decisions · Worksheet
The Heilmeier Catechism
The Heilmeier Catechism is a list of eight questions that DARPA, the research agency of the United States Department of Defense, asks about every proposed research program before the agency funds the program. George Heilmeier wrote the questions and ran DARPA from 1975 to 1977. The same eight questions work on any business proposal, from a new claims process to a software purchase. This page shows a weak and a strong answer to each question for one made-up proposal. At the end you answer the eight questions for your own proposal and print one page to take to the meeting.
Guess first
Where does this pitch fail?
A claims manager at a car insurer opens a funding request for $180,000 with this sentence:
“We will use AI to transform the claims experience with a digital-first platform.”
Tap an answer
Which of DARPA’s eight questions does the opening sentence fail first?
The eight questions
One proposal, asked eight ways
The proposal below is made up. The claims manager wants customers with minor car damage to send photos from their phone, so that an adjuster can estimate the repair without a visit. Tap each question to see DARPA’s wording, what the question tests, a weak answer and a strong answer.
Read the list with care
Eight questions, used to assess
DARPA publishes eight questions. Some guides list nine because they split “What are the risks?” into one question about risks and one about payoffs. Short videos about the catechism often show five of the questions and leave out the three about what is new, cost and time. Those three questions are where a proposal meets a budget, so keep all eight.
DARPA describes the catechism as the way the agency assesses and thinks through every proposed program. We know of no published figure for how many proposals fail the questions, so treat any claim of that kind with care. The questions were written for research programs. In a business proposal, the “exams” in question 8 are checkpoints with a number and a date.
The De-risking Ladder helps you strengthen the answer to question 5, and the MVP Slicer helps you design a small first test for question 8. If the proposal is a course for yourself, the course business case writes the request for your manager.
Your turn
Answer the eight questions for your proposal
Pick one proposal you will present or approve in the next month. Answer each question in one to three sentences, as you would say the answer out loud. The page checks each answer for words a listener would have to decode and for missing numbers, then prints your answers on one page. Your answers stay in this browser and are never sent to us.
Use it this week
Read question 1 aloud
Before your next funding meeting, read your answer to question 1 aloud to someone outside your team. Ask the listener to say back what will change and for whom. If the listener cannot, rewrite the answer before you touch questions 2 to 8. If you are the person approving the money, send the eight questions to the presenter a week before the meeting.
A proposal is ready for a decision when a stranger can say what will change, what the change costs, how long the change takes and how you will know the change worked.
Asking clear questions before a decision is one part of digital intelligence, and the 30-day course builds the rest of that skill one day at a time.
Sources
- DARPA, The Heilmeier Catechism (read 2026-11-04). The eight questions on this page are DARPA’s wording. The facts about George Heilmeier’s years at DARPA and the agency’s use of the questions come from the same page.
- A short video by the creator Steffen (@corecodevibes) prompted this page. The video lists nine questions and shows five of them, so we used DARPA’s list instead.
- The claims proposal, its figures, the weak and strong answers, the checks on your answers and the advice in “Use it this week” are our own.