At the risk of stating the obvious, AI took a leap in 2024. What was once a market of mostly better chatbots is now filled with tools that are transforming how organizations solve problems and deliver value.
Tools like v0 and Builder.io now enable teams to generate high-fidelity designs and code through a simple prompt, all in minutes. But these advances beg an important question: does building AI-enabled products fundamentally change the role of the product manager?
For the purposes of this post, we’ll focus on product managers incorporating AI features into existing products, rather than those building a new LLM.
The core role of a product manager
Before the AI floodgates opened, product managers across industries and teams focused on answering three key questions:
1. Who are our users?
- What are they trying to achieve?
- What problems or opportunities exist in our target space?
2. What can we do to solve their problems?
- What products, features, or improvements to the experience could add value?
- How can we validate our direction before investing significant resources?
3. Why should we be doing this?
- Is it the right thing to be working on, given our business goals, current capabilities, and opportunity costs?
There are a number of other ways to frame the above, such as Marty Cagan’s 4 Risks, but the core elements remain the same.
How AI changes the product manager’s mandate
1. Understanding your users
Understanding user behaviours, pain points, and opportunities remains as critical as ever. AI doesn’t change this fundamental requirement. Organizations must give product teams space to talk to users and explore opportunities before deciding how to incorporate AI into a given product.
However, AI is rapidly accelerating how product managers can synthesize and uncover insights from large swaths of qualitative and quantitative data. There are many dedicated tools in the market popping up in this space, but you start by simply dropping raw interview notes into ChatGPT and asking for some top line insights based on the context and opportunity space.
2. Defining how to solve problems
This is where AI brings significant changes compared to building a traditional product. AI introduces completely new ways to solve problems and deliver experiences:
- Scalability: AI enables teams to scale processes and functions previously bottlenecked by people. For example, in our work with Renolution, we leveraged AI to match contractors with jobs to create a human-in-the-loop system rather than relying solely on manual matching.
- Non-deterministic experiences: Product experiences are shifting from deterministic to non-deterministic experiences. This requires product managers to rethink the customer journey and how to provide customer support for products with these experiences.
Building AI-enabled products also requires addressing new considerations:
- Do we have team members who are knowledgeable in AI and can design and build with it?
- Do we have the necessary data for the model?
- How will we handle the data we receive?
While product managers don’t need to be developers, they do need a baseline understanding of AI technologies and relevant tech stacks to contribute meaningfully to the product’s design and strategic direction.
3. Balancing customer and business value
Finally, we have the why. Building excellent products requires balancing business and customer value. As the temptation to inject AI into everything or prioritize AI features just to check the box grows, product managers must remain outcome-driven to ensure what’s built delivers real value.
So what?
In short, building AI-enabled products doesn’t fundamentally change the product manager’s core mandate. There are undoubtedly new complexities and factors to consider when building products with AI, but just like with any new technology, AI is a means to an end. For product managers, that end is, as always, delivering value to both customers and the business.
