A useful experiment begins with a question the team cannot answer yet. Make the first test small enough to try, clear enough to discuss, and honest enough to change your mind.

SMALL TEAMS
LEARN FAST

Original thoughtful inventor with a yellow idea bulb and tiny rocket
Fieldwork Notes
A practical team guide
One question.
One useful next step.
Original three-person team around an idea bulb

FIND THE
QUESTION
BEFORE THE
BIG PLAN

Write down what must be true for the idea to work. Choose the assumption that matters most and that the team understands least. Describe the situation, the people involved and the behavior you want to understand. Avoid packing several unrelated uncertainties into one test.

A question gives the first test a purpose. It also makes the result easier to interpret. For example, ask where someone gets stuck in a task before deciding which feature to build. Keep the question visible while the team designs the activity.

A first test is a learning tool. Its scope should fit the question, the time available and the decision it can inform.
Agree on what the team will observe before beginning. Keep the plan simple enough that another person can explain it.

BUILD A
SMALL ENOUGH
FIRST STEP

Make one thing tangible. A sketch, conversation or rough model can help reveal what a longer plan might miss. Define a clear stopping point, and agree on who will record the observations. The first version only needs enough detail to make the question testable.

Original teammates testing a paper bridge
A result becomes more useful when the team knows what it was looking for and what the test could not show.

LOOK FOR
USEFUL
SIGNALS,
NOT APPLAUSE

Original teammates around a launch experiment

Notice what people do, where they hesitate and what they ask without prompting. Keep the observation separate from your interpretation. Record a concrete action or phrase before writing down what you think it means. This gives the team something specific to revisit together.

Encouraging comments can start a conversation. They do not, by themselves, answer every question about an idea. Look for the difference between liking a concept and being able to use it. When the evidence is ambiguous, describe the uncertainty instead of forcing a conclusion.

USE TOOLS TO
SHORTEN THE
LEARNING LOOP

Pick the simplest tool that makes the question easier to explore. Add complexity only when the test needs it.

A shared sketch can expose different assumptions. A small prototype can make an interaction visible. A clear record can help the team compare what happened.

The useful output is a better-informed decision. Before adding another tool, ask what it will help the team notice or explain. A more elaborate setup is only worthwhile when it improves the quality of the learning.

Original cloud and network learning scene with two adults

MAKE LEARNING
A SHARED
TEAM HABIT

Give each person a clear part in the experiment: ask, observe, make or record. Leave room for different interpretations.

End with a short conversation about what changed and what the team needs to understand next. Invite the observer and the maker to compare their notes. Write down the disagreement as clearly as the shared conclusion, then choose one next action together.

Original person climbing outlined steps beside a yellow arrow

INVITE FEEDBACK
THAT CHANGES
THE WORK

Original woman comparing two abstract feedback cards

Ask about a recent experience, a concrete difficulty or the way someone would approach the task. Let the person finish before explaining your idea.

Keep useful surprises visible, especially when they challenge the plan. Ask for an example when a comment is broad. A specific story about a recent experience often gives the team more to work with than a general opinion about a future idea.

A good follow-up question helps clarify what happened. It does not steer the person toward the answer the team wants.
An unexpected result is a reason to inspect the test and the assumption. It is not automatically proof that the whole idea is wrong.
Ask what was observed, what may have influenced it and which part of the question remains unresolved.

LEARN FROM
THE UNEXPECTED

Keep the useful evidence. Revise the explanation. Choose a next step that reflects what the team now understands. Separate a weak assumption from a poorly designed test, and note which part you would change before trying again.

Original two-person reflection scene at a winding-path whiteboard

KEEP THE
DECISION
CONNECTED TO
THE QUESTION

Original kneeling facilitator with a feedback-loop board
Return to the assumption you started with. Describe what the test supports, what it challenges and what it cannot resolve.

The next step may be to continue, change the approach or stop. Make the reason visible so another teammate can understand it.

A useful decision is proportionate to the evidence available. Record the reason, the remaining uncertainty and the person responsible for the next step. Revisit the decision when new information changes the picture.

MAKE ONE
TEST CONCRETE

Original two-person paper-bridge experiment
01State the question in one sentence.
02Choose a small action that can reveal something.
03Record what happened before interpreting it.

COMPARE THE
USEFUL DIFFERENCES

Change one meaningful part of the test when possible, and keep a record of what changed.
Original person examining three small test objects
A comparison is easier to discuss when the team can explain how the conditions differ.

START SMALL.
LEARN CLEARLY.

Original team celebrating a small shared experiment
Choose one question worth answering.

Give the next step a clear purpose.
Make room for what the result can teach you.
Fieldwork Notes / A practical team guide
·