Product managers aren’t doomed by the emergence of AI, says OpenAI product lead Britt Jamison. Quite the opposite – as AI takes over rote product management tasks, PMs will be promoted to a more strategic role.

60% of product managers already use AI daily in their work – so if you’re not harnessing it yet, now’s the time to start. Here are Britt’s 4 steps for integrating AI into your product work.

4 steps to integrate AI into product work

1. Find a workflow that saps your time or energy

Start by finding a workflow that involves a lot of time, brainpower, or creativity.

Think about workflows that:

• You complete frequently
• Takes you a lot of time (and that time never feels well-spent) OR
• Takes a lot of your mental energy OR
• Requires a ton of creativity

Here are some examples from Britt that sap his time, energy, or creativity.

2. Map those workflows against AI’s strengths

AI can’t do everything well – so make sure to choose workflows it can excel at. Here are a few examples:

• Research: Instead of scouring the internet for the information you need, ask ChatGPT to summarize it. It can help you more easily source industry reports and compliance or competitor research.some text
  • Go deeper: Ask follow-up questions on the research it finds, such as how to implement what it shared or what pitfalls you should anticipate.

• Ideation: AI can help you expand your thinking by suggesting initial ideas and offering suggestions on your thoughts. Some great use cases include product problem solving, scenario simulation, and providing feedback on pitches.some text
  • Go deeper: Ask follow up questions on the ideas you came up with together, such as which techniques make the most sense for your specific product and how to communicate a decision internally.

• Creation: Cut down on tedious manual tasks such as generating product requirements documents, developing users personas, and creating code by asking AI to do it for you. some text
  • Go deeper: Ask AI to supplement what it created with collateral, such as visual elements or a communications plan to help present what you created together.

• Synthesis: Use AI to analyze past project results and forecast future performance. Or have it combine notes from a cross-functional collaboration to determine where there may be conflicts or dependencies to consider.some text
  • Go deeper: Ask AI to round out its findings with third-party data or shorten what it came up with to make it quicker for executives to scan. Then have it roleplay as the executive and ask it where you can expect pushback.

Step 3: Execute on a workflow

Once you’ve chosen a workflow you believe AI can augment, it’s time to prove yourself right. Here are some implementation principles to follow while testing out the effectiveness of its assistance:

1. Clearly define what you want to get out of it – This could be higher-quality work, more creative approaches to tasks, or just getting things done faster.

2. Decide how quantitative you want to be – Your KPIs don’t need to be an exact science. You could choose metrics such as decision-making speed, hours saved this week, or even job satisfaction.

3. Check in weekly on your progress – Don’t wait 30 days to see if it worked. Review how things are going on a weekly basis to see if your efforts are moving the needle at all.

Step 4: Record your learnings (and retest!)

Data and metrics are great for understanding how (un)successfully you + AI completed a workflow, but you’ll get more out of this experience if you tell the full story. So combine the things you tracked in your weekly check-ins with the things that you learned throughout the process.

Once you can take a look at the whole picture, it’s time to decide if you want to hire AI full-time to take on this task or if you need to tweak your hypothesis.

Are you a Section member yet? Members have access to all of our courses taught by industry leaders on everything from AI to product strategy. Take advantage of our limited time pricing and start learning.