You will be able to present your results in a short note that supports decisions about tools and ways of working.
Your results note goes to the person who can act on it, usually your manager, and it should fit in about half a page. Put the measured result in the first two sentences, explain your method briefly, say which risks you managed, and finish with one request your manager can approve.
You've already done the measuring. What's left is writing it up so that your manager does something with it.
The most common mistake is to open with how you approached the course: the modules you worked through and what you learned about prompting. The results and the request then come late in the note. Managers who read email on their phone between other duties may stop reading early, so the important content has to come first.
Priya, an example employee, measured how much time AI saved her on her own tasks and wrote to her regional manager about it. Her first draft spent three paragraphs on how she approached the course, the modules she worked through and what she learned about prompting. Her results came in paragraph four. Her request, for team access to the company's business assistant, was in the last line. Her manager reads email on her phone between site visits and would probably have stopped reading after paragraph two.
The first rule is to lead with the result. In the first two sentences, say what changed, in numbers: which tasks saved time, how much time per week, and what you did with the time you saved. Everything else in the note supports those opening sentences.
Priya rewrote her opening. It said that over six weeks of testing, using AI on three of her regular tasks saved about two hours twenty minutes a week, measured against a two-week baseline. It also said she had used most of that time to run supplier performance reviews that had been postponed since March.
People often leave out what they did with the saved time, but it is often what a manager cares about most. "Two hours a week saved" is abstract. Two hours a week spent on supplier reviews that had been slipping is value your manager can see. If the saved time went into more of the same work, say so, because that still counts as a result.
Managers have heard many claims about AI saving time, and they are right to be sceptical. A sentence or two on method makes your figures believable. It's usually enough to say three things: you timed the tasks before changing anything, you timed the new method from start to finish including checking, and you used averages rather than your best runs.
If something weakens your figures, such as a baseline that was partly an estimate, say so yourself. That is far better than having someone else discover it. Offer to send your detailed log if your manager wants it, but keep the log out of the note.
Many managers worry less about whether AI saves time and more about what could go wrong. A sentence on the risks you handled answers that worry before your manager raises it. Three kinds of risk are worth mentioning:
the data rules you followed, such as using only the approved tool and redacting supplier details before drafting errors your checks caught, such as a check cell that found a miscoded order, or summary items you corrected before sending any task you decided to keep doing yourself, and why
Priya's note said she used only the company's approved tool for work data and redacted supplier names before drafting. It also said her report's check cell caught an order her old method had been missing. She added that she had stopped using AI for the monthly regional update, because checking it took longer than writing it herself. Each of these shows judgement, which is what a manager wants to see before encouraging others to use AI.
If your note has no request, the usual reply is "thanks, great work" and nothing changes. Your request needs to be concrete and specific enough for your manager to say yes to it. "We should use AI more" is too vague to approve. Priya asked for access to the company's business assistant for the three supervisors in her team, so she could share her email and meeting templates with them. Other good requests include time to run a short session for colleagues, approval for a tool that isn't yet on the approved list, or a decision on a question the AI policy doesn't answer.
Make only one request. A manager can say yes to one thing in a reply, but several requests go into a pile to think about later.
Priya's final note was about 200 words. It gave the result and what she did with the time, then two sentences on method, three on risks managed, and the request for her supervisors' access. Her manager replied the same day and asked IT to set up the access.
Write your note in the same order. Open with the result in numbers, add a sentence or two on method and anything that weakens it, cover the risks you managed, and end with your one request. Keep the whole note to about half a page.
Before you start your own note, put your figures from lesson 9.1 and your list of risks handled side by side, and decide on the one request you will make.
Write a half-page note to your manager with your results, method, risks handled and one specific request.
Junxiong-WFG Organisation is an authorised representative of AIA Financial Advisers Private Limited (Reg. No. 201715016G).