You’ve likely seen countless articles on why AI is transformative, and you’ve read about the various and sundry LLMs that are out there.

This article takes a different approach: how do you actually get your product management team to consistently use AI to help level up your team?

While I don’t claim to have cracked this nut entirely, it is something that I’ve worked on, and thought it would be helpful to share with other product people interested in bringing AI to their teams.

The process has looked something like this:

Make sure you understand your company’s policies around LLM/AI usage

• This likely incorporates details around data security, such as not including PII even in secured paid accounts, and nothing company-specific in any free LLMs

Get access to any paid AI accounts available at your company, and start using the AI(s) yourself

• You have to understand the possibilities yourself first

Add a few eager early adopters to the AIs and encourage their usage

• Ensure they understand the company LLM//AI policy before they dive in

After a few weeks of this, bring the team together:

• Have them read a relevant article before the meeting if possible. We read Lenny’s How to Use Perplexity in Your PM Work
  • While many had been interested in AI, this helped drive an understanding of practical uses
• Review the company policy around LLM/AI policy
  • Not only is this important for the company, it’s important for your team: if they don’t feel comfortable knowing what they can and can’t put in which tools, they will be less likely to use them
• Encourage the early adopters to share their use cases, including screensharing where possible
  • Share any useful company-specific prompts that people have created with the team
    • For example, I created a prompt to write story tickets and another to write bug tickets using our standard format
  • The early adopters shared other use cases as well, such as using LLMs to assist with:
    • SQL queries
    • Email and help article copywriting
    • Qualitative data analysis
  • We also shared where AI has not proven useful
    • I shared where I’d tried to use it for something math-related, and it totally fell on its face. Luckily, I realized this just before sharing inaccurate data with all of the leadership team!
• Run a mini-workshop
  • We did this with a free version of Perplexity, where I gave my team a particular item to research related to decreasing churn within their respective products, and they each researched it with that tool instead of using Google. Then several of them shared what they found.
• If there are multiple tools the team can use, make sure it’s clear which are best for which use cases

Once the team has access to the tools, ask the team on a regular basis what they are using AI for

• For example, one team member brought up that ChatGPT enables you to install add-on modules, and she’d used one of those to create a complex flow diagram that would have otherwise taken her a lot of time
• I shared with the team that I realized you could upload images to ChatGPT and it could ‘read’ the image, speeding up my ticket writing time when I wanted the copy from designs in the ticket

Share additional AI resources with the team, and encourage them to do the same

• I’m a fan of the AI for Work newsletter, and encouraged my team to sign up as a source of ongoing AI news and information

I hope you find these ideas helpful in bringing AI to your own teams. Best of luck!