Ask about friction before choosing a tool

People who have done the work for years know where exceptions appear. Ask which recurring step causes retyping, searching, or avoidable delay. Also ask which parts require relationship decisions or a formal approval. The first list may suggest a test; the second sets its limits.

Make the test easy to reverse

A first test might write an internal agenda from approved notes or sort public questions for a team member to review. Keep the original process available. Do not start with automatic customer replies, record changes, or private employee material.

Write down who owns the test, which tool is permitted, what information may be entered, who reviews each output, and who can stop the trial.

Measure the whole task

Record preparation, asking AI, checking, correction, and transfer time for several attempts. Note errors that were caught and errors that would have reached someone else. If the team spends more time cleaning up than the old process took, change the task or stop the test.

Ask the people doing the work whether the answer helps. A manager seeing a polished sample is not the same as a team finding a reliable routine.

Try it yourself

Write a one-task pilot brief

Discuss a routine task with the person who currently does it and confirm workplace AI policy before using any material.

Before you paste: use made-up, public, or approved information. Before using work information, check that your workplace permits both the tool and the information.

Instructions to complete before you use them

Replace any text in square brackets. Read the complete instructions once before using them.

Write a one-page pilot brief from these approved notes only. Include the current task, proposed AI step, permitted input, human reviewer, stop rule, measures for time and errors, and a date to decide whether to continue. Put missing decisions under Open questions. Do not assume tool approval.

You have finished when you can check these points

  • Did the people doing the task help define a good result?
  • Can the trial stop without disrupting the existing process?
  • Will you count checking and correction time?

Keep these points

  • Start with a team problem, not a product demo.
  • Give the test a named owner and a stop rule.
  • Assess the full method, not the starting version alone.

Continue for free

Try another task or ask a question.

Get a complete practical task each week in the free newsletter. If you are stuck, ask in the free Skool community without sharing private details.

Optional paid help with Ikram Rana

Get help with a recurring work task.

Bring a task you already understand and the tools your workplace permits. Discuss where AI could help and what you need to review before using the result.

Book a call about your task

Agree on what is included and what it costs before work starts. The guides stay free.