Leveraging AI as a Product Manager

TL;DR: Generative AI, LLMs, and Predictive engines are transforming ways of working beyond engineering. Professionals across a diverse set of domains are utilizing AI to improve the efficiency of day-to-day workflows on an individual level. Product management has extensive potential for AI disruption that can improve the efficiency and quality of product delivery. The blog talks about major product management use cases for AI.

There’s a growing sentiment in the tech community that, ‘product management is dead … or nearly there’, or ‘the era of hiring product managers or building a product management unit in the orgs is no longer important’. A large chunk of this sentiment is thanks to the quickly growing popularity of generative AI tools like Gemini, ChatGPT, Claude, Midjourney, and more. Because why do you need a product manager if you can feed facts to AI and ask it to design a feature, summarize feedback, or build a wireframe, right?

Quick Recap of a (Dying?) Role

The idea of a “Product Manager” is inspired by Neil McElroy’s “brand-man”, who thought that people who can combine the knowledge of brand and production are needed to grow Procter & Gamble. The IT industry, particularly Hewlett-Packard in the 1970s, took this idea and made innovation part of the job description as well. In the 1990s, the software industry started using the term “product people” inspired by brand-man who now also wear marketing hats but focus on developing digital products. Microsoft, Apple, and Google are some prime examples of those companies. At the turn of the millennium the product people, now called product managers became agile. 

Today, product managers have strategic roles in organizations to build successful products. 

An interesting take on product people by Steve Jobs, considered the greatest product manager ever:
 

So, is product management really dead?

The domain has always evolved with the changing landscape of new markets, technologies, and business needs. The first product manager was a combination of sales, marketing, and brand. It evolved to absorb expertise like R&D, user experience, data science, and project management.  As AI models and technologies strengthen, naturally they impact roles. Did online learning kill academia? No. But what it did was; improve ways of learning and make education more accessible. 

Saying product managers need to learn about AI is no different than suggesting, almost a decade ago, that PMs should learn about Firebase, Figma, APIs, or Cloud.

AI-driven product management is a great opportunity for professionals to be more impactful. We will dive into ways to use AI in your daily job. To summarize here’s what product managers normally do in big or small organizations:
 

1- Develop product vision and strategy

2- Manage conversations with stakeholders

3- Collect customer Insights and market research

4- Plan development and prioritize needs

5- Collaborate with other functions

6- Manage launch and go-to-market strategy

7- Optimize and scale products

How Can You Leverage AI to Do Your Job Better More Efficiently?

Note: A lot of these techniques are either based on my personal experience or shared by the wider product community. Go to the end of this article to submit questions or contribute more insights. 

Meetings and Conversations

Take a look at your calendar. Are you overwhelmed? I am going to go out on a limb and say yes. For PMs, a significant portion of the day is spent on talking to people. Speaking to users, giving feedback to customer support, answering engineers, going back and forth with sales - and so much more. It’s not just attending meetings that is overwhelming but the pre and post rituals also take up significant time. Now there are tools to help you do that without spending hours on mundane tasks.

Event Scheduling: Stop following up, scheduling, and rescheduling again and again with team members in different time zones, or across 10+ calendars. Use AI assistants to help you do this on your behalf like SuperCal or Trevor AI.

Note-Taking: As PMs, most of the time attending is not the only requirement. Meetings are where most of the important decisions are being made, and expectations are set. So you can be more attentive to the discussion itself, use assistants like Spinach.io or TL;DV to take notes on your behalf and share them with participants automatically.

Action Items: When it's time crunch or there’s increased work on your plate, use assistants like Fireflies or Notta to follow up on action items from the meeting independently without sending multiple follow-ups. These tools are also integrated with communication tools like Slack, Teams, and Gmail.

Documentations

Do you have multiple asks for requirements, answers, or directions from cross-functional teams to refer to in different product capacities? Finding time to dedicatedly write content, structure it, and disseminate it can be dog work. Replace processes or steps that do not require independent thinking or ideation with generative AI assistance. 

Requirements: Have a conversation with Chatprd.ai or Jasper about what, why, when, and how of what you want to build without focusing more on the content outline, and the AI bot will convert your thoughts into structured tickets, PRDs, or FSDs. 

Diagrams: Generate system or user diagrams and user flows or mindmaps using the Edrawmax tool for quicker conversations and expectations setting with engineering teams.

Presentations: Feed basic prompts or use existing libraries to generate slides with easy-to-grab visuals and images using Beautiful.ai. Doing so will help reduce the burden of designing visuals and buy more time to improve the quality of content.

Strategy and Prioritization

Though AI cannot replace human decision-making, it can help ideate conditions or suggest frameworks to capture outcomes of strategic discussions and research. Here’s how:

Roadmap: Use Clickup or Roadmap.sh to generate structured roadmaps based on attributes like themes, timelines, metrics, and more. You are able to prompt AI engines to connect business goals versus customer needs to develop a plan or roadmap for engineering, marketing, product, etc. The tools give you a granular or collective view of roadmaps depending on the set of stakeholders you are talking to. 

Prioritization: Clickup AI and Zeda.io can immensely help with categorizing customer and stakeholder feedback, converting insights into ideas, and quick baseline prioritization of needs and wants. You can define attributes, levers, or metrics as input and get prioritized lists and automated empathy mapping as output. This will allow you to view quick iterations and modify them without spending too much time on steps that don't require intelligent decision-making.

Research

In an era where new products and features are being launched every second, and data makes it easier to track trends, the competitive landscape has become even more fierce. R&D needs significant time not only to find information but also to absorb it for decision-making.

Market Insights: Summarize quantifiable insights from research papers with Qualtrics, or ChatGPT. Use intelligent prompts to identify and summarize specific information or insights like quantitative data, problems, solutions, trends, etc. without sifting through huge research papers.

Competitive Analysis: Generate frameworks to capture information and analyze it for finding opportunities using Crayon or Competely. These tools can read through online or prompted resources to summarize insights in the form of competitive analysis. This makes it easier to keep track of the market and competitors and have information at the ready in a digestible format.

Design and Testing

Ever get the feeling that you simply aren’t able to test 5 different ideas or experiences because of how time-consuming it can be to visualize them? So you end up doing mental gymnastics to build two iterations and that too very basic. Sometimes with more complex use cases or features, it’s difficult to collect helpful feedback by using simple design prototypes for testing.

Wireframing/User Interface: Design different iterations of basic user experiences to compare feasibility or for the quick show and tells with a team with the help of Uizard and Galileo. These tools especially help you if you don't have a large UX team at your disposal.

Operational Prototypes: Build and deploy fully developed MVPs or prototypes to run real-life user experiments without sacrificing your engineering quota with tools like Replit.ai and Codepen. You can also define the tech stack in case you plan to use MVP as a baseline for development. These tools require a basic understanding of coding languages and database design.
 

Considering AI as a New Process Skill to Be Acquired

There is no denying that AI is revolutionizing roles across domains. For product folks: it can automate routine tasks, improve customer and stakeholder insights, and speed up decision-making.

But, it's not ready to replace a product manager’s creativity, instinct, and empathy - because it simply cannot think on its own without relying on trillions of human-produced content. As new trends, needs, and markets emerge, you need a human perspective to make sense of noise and uncover opportunities.

Product managers however can and should leverage AI to streamline manual processes that do not require much thought or creativity, improve product development by adding more tools, and deliver better solutions more frequently. To fully realize the value of generative or predictive AI, sharp product managers can combine new-gen technologies with product thinking. 

Parting Thoughts

There's a lot out there to unlearn, learn, or relearn, especially as a product manager. The expectation should not be to learn everything all at once but rather learn one thing, use it to make an impact, and then expand your knowledge further.

How did I use AI to empower myself to write this article?

• I asked Gemini to suggest an outline based on what message I wanted to deliver based on my writing style.
• I used Venngage to convert text into an infographic.
• I asked Gemini some factual questions instead of sifting through multiple searches and resources.

Previously a well-researched, and articulated article would take me easily 6-10 hours (more if you count distractions). But with AI I was able to cut that number by half without sacrificing the quality of written text.

There is an evolving list of technical and behavioral competencies that make up a successful product manager. AI has a huge potential to improve your efficiency but it is not the definite answer. The goal of utilizing AI is to improve productivity or efficiency, and not use it for the sake of using it.

If you are a product manager who is currently leveraging AI in their daily routine then write to us at marketing@arbisoft.com if you wish to contribute more suggestions for this article.