Skip to content
Apply now
← Field Notes

Grow

What to do when a project doesn’t work

A failed test is not a verdict on the student or the idea; it is information that can reveal a weak assumption, unsuitable method or scope that needs to change.

Boundlearn Editorial

A digital project checklist being reviewed and revised

Projects are often presented after uncertainty has been edited away. The public version shows the final prototype, argument or event, not the test that produced no response or the method that had to be abandoned. Students can therefore interpret ordinary failure as evidence that they do not belong in difficult work. A more useful response is diagnostic. Separate the result from identity, identify what the attempt has revealed and choose a next action that is meaningfully different. Failure becomes educational when it changes judgment.

Describe the failure precisely

'It did not work' can mean users ignored it, the code crashed, the evidence contradicted the hypothesis, the team missed deadlines or the student lost interest. Name the observable result without turning it into a judgment about ability. Precision makes the problem smaller and identifies whether the next step belongs in the idea, method, execution, collaboration or commitment.

Replace ‘nobody liked it’ with an observation you can check: how many people saw the idea, what they were asked to do and where they stopped. A small test may leave important questions unanswered.

Return to the assumption

Every project contains beliefs about people, materials and cause: users have this problem, this measurement represents the outcome, this feature will help, this source is reliable. Identify which assumption the result has challenged. Sometimes execution was weak; sometimes the project has produced important evidence that the original premise was wrong. Do not repair presentation before understanding that distinction.

Check the method

A test may fail to answer the question because the sample is unsuitable, instructions are ambiguous or too many variables changed at once. Ask whether another person could repeat the process and understand what was measured. Improve the method before drawing a dramatic conclusion. This is especially important in research, where an exciting negative result may simply reflect a design that cannot support the claim.

Check whether the instructions, materials or measurements were consistent. If they were not, a revised test may tell you more than changing the whole idea immediately.

Reduce the scope

Projects often stop moving because the next step contains several unknowns at once. Narrow the audience, shorten the time period, test one feature or analyse one case. A smaller question can restore the connection between action and evidence. Scope reduction is not giving up on ambition; it is creating a version that can teach the student what the larger idea requires.

Decide whether to persist, pivot or stop

Persistence is valuable only when the next attempt has a reason to differ. Continue when new evidence or knowledge suggests a better test. Pivot when the original problem remains meaningful but the proposed response does not. Stop when the purpose has disappeared, the costs are disproportionate or the key question has been answered. Each choice can demonstrate judgment.

Set a limit before the next attempt: a question to answer, the time available and what would make stopping reasonable. Discuss safety or ethical concerns with a qualified adult before proceeding.

Document the honest outcome

Record the unsuccessful result, the diagnosis and the decision that followed. Do not hide early drafts or rewrite the story as if the final direction was obvious from the beginning. The trail can become a method note, reflection, portfolio annotation or handover for someone else. Honest documentation turns difficulty into transferable learning without pretending that failure itself is automatically valuable.

What to carry forward

The aim is not to celebrate every failure or persist indefinitely. It is to respond with accuracy. A project that ends after a careful test can be more complete than one kept alive through cosmetic changes. Ask what is now known, which uncertainty remains and whether another attempt is justified. The quality of that decision—not the appearance of uninterrupted success—is the capability worth carrying forward.

Sources and further reading

This essay draws on Boundlearn's editorial framework for purposeful, mentored learning.