An AI plan helps your team agree on what to improve, where AI might help, and how you will know whether it worked.
You can start with one page. Use Think. Plan. Act. to work through the questions together.
Think: what needs to improve?
Talk to the people doing the work. What takes longer than it should? Where do mistakes happen? What do customers or colleagues keep asking for?
Choose a clear goal. “Use AI in sales” is hard to act on. “Help our team prepare customer proposals with fewer mistakes and less rewriting” gives you something to work toward.
Ask someone to walk you through a recent proposal. Where did the information come from? Who wrote it? Who checked it? What had to be corrected, and why?
You may find that the team needs clearer instructions, better information or some training. Put those changes in the plan too. AI may help with part of the work, but it will not settle a question your team has not answered.
Plan: decide what to try and who will lead it
Describe how you want the work to happen. Then decide which part AI could help with and which decisions people will make.
For example, imagine a team that writes customer proposals. It could start by agreeing on the questions to ask each customer and keeping its service descriptions up to date. AI could use that information to draft an outline. The team would still decide prices, agree on delivery dates and approve the proposal.
Choose one person to lead a small trial. Agree on the task, what information the team may share with the tool, who will check the answers, and how much time and money to spend. Give the people taking part time to learn and tell you what is working.
Now choose a tool that fits the job and your team's rules for using business information. A conversation in an approved AI tool may be enough for the first trial. You can decide later whether it is worth connecting it to other systems.
Act: try it and compare the results
Before the trial, look at a few examples of how the work gets done now. Record the time spent preparing and checking proposals, the corrections needed, and any customer requests that were missed. Include difficult examples as well as easy ones.
Use the same checks during the trial. Count the time spent fixing AI's answers. If writing gets faster but checking takes much longer, you may not be saving time. Fewer mistakes or clearer proposals can also make the work worthwhile.
Agree before you start on what would make the trial worth continuing. At the end, decide whether to keep going, change the approach or stop. If nobody can check the answers well enough, solve that problem before giving AI more responsibility.
Keep a short record of what you learned. Use it when you choose the next task or try another tool.
Write your one-page plan
Fill in these sentences with your team:
- Our goal: We want to make ___ easier or better for ___.
- The problem: Today, we do the work by ___. The hardest part is ___.
- The change: We will improve ___ first. AI could help us with ___.
- The people: ___ will lead the trial, and ___ will check the answers.
- The rules: We may share ___ with the tool. A person must decide or approve ___.
- The trial: We will try ___ tasks over ___. We can spend ___ hours and ___ dollars.
- The check: We will compare ___ with how we do the work now. Every result must ___.
- The next decision: On ___, we will review what happened and decide whether to continue, change or stop.
If you cannot fill in a sentence yet, write down what you need to find out and who can help.