Leveraging AI in Product Management
Understanding how to work smarter, not harder in the AI age.

Work Smarter, Not Harder
Product managers are often juggling conflicting priorities.
The good news? The creativity of product and technology experts just like you are working hard to create the next great time saver to handle our hectic schedules and free us to concentrate on what truly important: value creation for our clients.
Understanding What It Does vs. the Tools
The past few years have had rapid innovation in the realm of consumer generative AI tooling, automation, and virtual assistants. It’s now common to find tailored technologies appropriate for our day-to-day processes integrated into the tools we already use or available for free.
Investigate and Discovery
These days, artificial intelligence performs the heavy work of tracking rival actions, cross-channel customer sentiment analysis, and early trend detection before they are clear-cut. These techniques surface insights that would take days to manually uncover, not only gather facts.
How do you get started?
- Consider research-focused tools like Perplexity and Phind that support not only searching the web, but specialize in creating composite responses from multiple data sources.
Automation of Workflows
Remember when we spent hours organizing material, following progress updates, and transcribing meetings? Now handling these chores, AI helpers automatically create specs, find action items from meetings, and keep everyone in sync without continual pestering.
ProductPlan’s 2024 Product Management Report and validated by McKinsey’s research on PM workflow optimization shows that AI solutions enable PMs to quadruple time spent on high-value activities while lowering administrative burden by 75%. What did those product managers instead spend their time on? Value-add, human-centric tasks like stakeholder engagement and strategic thinking.
How do you get started?
- Most meeting tools already include AI–whether you’re using Zoom, Teams, Google Meet, or Slack, there’s likely an AI recording and transcription tool available.
- Taking meeting tools a step further. Most meetings get “recorded,” but who has time to watch hours of recordings? These AI tools generate not only the transcript, but can generate task lists, TODOs, and full meeting notes so you can dig in. Some systems, like Teams, actually let you “chat” with the meeting to ask questions and see if your questions were answered in the meeting.
- If you’re looking for a bit of hardware and have the budget, check out great devices over at PLAUD.
Analytics and Support for Decisions
Instead of drowning in dashboards, we now have artificial intelligence tools that aggressively flag significant trends in user behavior, assess feature proposals against strategic goals, and even simulate the possible impact of changes before we commit resources.
How do you get started?
- ChatGPT, Phind, Claude, and many others support data uploads and analysis. Keep in mind that your data security is vital, but aggregate, anonimitized data can easily be analyzied, trends found, anomalies detected and more with these tools.
- If security is your primary concern, consider creating a localized model with a tool such as Sanctum which runs many open-source LLMs locally and encrypted on your machine.
Actual Advantages, Not Just Buzz
Our daily work is clearly impacted:
Most PMs say they save 10 to 15 hours every week on chores AI now handles—that is almost two full workdays to concentrate on strategy and creativity. AI creates emerging insights from more data than we could personally handle speeds us toward better judgments free from analysis paralysis.
Beginning small and thinking big
The most successful PMs avoid trying to change everything over night. rather, they:
- Starting with clear time-sinks: With very little risk, meeting summaries, documentation updates, and competitor tracking provide instant relief.
- Set reasonable expectations; these instruments are tools, not substitutes; they will still make sporadic errors needing your judgment.
- As you become comfortable with simple technologies, progressively investigate more strategic uses such impact modeling and prioritizing help.
What am I using it for?
As a developer at heart, I’m naturally lazy and prone to spend a day writing a script to automate a 10 second process so I never have to do it again. AI’s automation, research capabilities, and virtual assistants act as my “squad of 1” support group.
- Ideation - using a pseudo RAG to bounce ideas around, red team theories and assumptions, and provide a sounding board.
- Automation - review feature concepts and storyboards and build out skeletons for features, user stories, acceptance criteria, and help me find gaps.
- Research - reviewing key trends and customer sentiment in various markets on social media.
- Development - rapidly prototyping object models, database structures, and rough UIs to quickly stand up proof of concepts and low fidelity prototypes in minutes.
- Education - cataloguing books, research papers, white papers, and opinion pieces to quickly get a view on new topics in technology, product, health, or whatever else interests me.
The list could likely go on as I find a new way to use AI every day.
The Final Thought
Fear not that AI will take your job, but that someone empowered by AI will who is working faster and smarter than you.
These tools free us to concentrate on the uniquely human elements of product management: knowing customer demands, creating appealing visions, and negotiating the difficult human dynamics of product development by unloading the ordinary chores that occupy our days.
Share this post
Twitter
Facebook
Reddit
LinkedIn
Pinterest
Email