← Back to blog

I do not know where to start: five ways to find the first step

Five ways to find the nearest action in an unclear task: work backward, identify a question, inspect an example, run a test, or ask.

“Start the project” is not an action. It forces you to choose a direction, evaluate risk, and perform the work simultaneously. Uncertainty easily becomes delay.

Five ways to find an entry point

  1. Work backward from the result. What must exist immediately before final delivery? What must exist before that?
  2. Find the central question. Which one answer would change the plan most?
  3. Inspect an example. Use it to see the shape of the result, not to copy it.
  4. Run a small test. Draft one paragraph, calculate ten rows, or mock up one screen.
  5. Ask the recipient. Clarify the criterion, deadline, or preferred format.

Check the step

A useful first step:

  • begins with a verb;
  • fits inside a limited time;
  • uses a specific object;
  • ends with an observable change.

“Research the market” fails the test. “Record the prices of five competitors in a table” passes.

When uncertainty is high, the first step may only reduce uncertainty. A sent question changes the situation even while the document remains empty.

Leave the second step immediately after finishing. The next start will not require another reconstruction.

For a vague assignment, use how to clarify an unclear task. For projects with several unknowns, see starting with many unknowns.

Frequently asked questions

What makes a good first step?

It is observable, limited, requires no new decision, and creates information or material for the next action.

Do I need the entire plan first?

No. Under high uncertainty, the first step may be a question, experiment, or short investigation that makes the plan more accurate.

What if the first step is still too large?

Ask what action must happen immediately before it and repeat until you reach a step that fits in 5–15 minutes.