The Google Way to Use Machine Learning for PMs

Aviral Vaid

Hatched by Aviral Vaid

Aug 12, 2023

4 min read


The Google Way to Use Machine Learning for PMs

In today's fast-paced and data-driven world, product managers (PMs) are constantly faced with the challenge of finding the best solutions to complex problems. This is where machine learning (ML) comes into play. ML is a powerful tool that can help PMs make sense of large amounts of data and uncover patterns that might not be immediately apparent. But how exactly can PMs leverage ML to their advantage?

The key to using ML effectively is to have both the data and the answers. In other words, PMs need to have a clear understanding of what they want to achieve and the data that is available to them. This is because ML is all about figuring out the underlying rules and patterns that govern a particular problem. By feeding the data into an ML algorithm, PMs can let the machine do the heavy lifting of finding these rules and patterns for them.

But what if the problem at hand is not well-defined and there are no clear rules to follow? This is where the opportunity solution tree comes in. Ericsson, a renowned expert in the field, argues that better mental representations are what set experts apart from novices. He believes that by asking the right questions, PMs can uncover multiple potential solutions to a problem.

During a recent conference, Ericsson asked attendees to identify something they wanted in their life. He gave examples such as a house, a better job, or more leisure time. After giving the audience a minute to write down their answers, he posed another question: "If you had whatever you wrote down today, what would that do for you?" This simple yet powerful question forced the audience to think beyond their initial solution and consider the underlying benefits they were seeking.

For example, if someone wrote down "owning a house," Ericsson would then ask, "How else might you feel more grounded in your community?" This question challenged the audience to think expansively and explore alternative ways of achieving the same goal. By reframing the problem in this way, PMs can uncover a multitude of potential solutions that they may not have considered before.

This concept aligns with Jonassen's argument that ill-structured problems have many solutions, with no right or wrong answers, only better or worse ones. To effectively tackle these problems, PMs must first define the desired outcome and constraints before exploring potential solutions. This approach, known as good product discovery, ensures that PMs have a clear understanding of what they want to achieve before diving into the problem-solving process.

So, how can PMs apply these insights and leverage ML to their advantage? Here are three actionable pieces of advice:

  • 1. Start with a clear desired outcome: Before diving into the problem-solving process, PMs should have a clear understanding of what they want to achieve. By defining the desired outcome, PMs can guide the ML algorithm towards finding the right patterns and rules that will help them achieve their goal.
  • 2. Embrace ill-structured problems: Instead of searching for a single solution, PMs should embrace the idea that there are multiple ways to solve a problem. By reframing the problem and asking powerful questions, PMs can uncover alternative solutions that they may not have considered before. This can lead to innovative and unexpected outcomes.
  • 3. Experiment and iterate: ML is not a one-time solution. It requires continuous experimentation and iteration to refine the results. PMs should be willing to try different approaches, test various hypotheses, and gather feedback from users. This iterative process will help PMs fine-tune their ML models and improve the overall effectiveness of their solutions.

In conclusion, ML can be a powerful tool for PMs when faced with complex problems. By leveraging ML algorithms, PMs can uncover hidden patterns and rules that may not be immediately apparent. Additionally, by embracing the opportunity solution tree concept and asking powerful questions, PMs can uncover multiple potential solutions to ill-structured problems. By following these actionable advice and continuously iterating on their ML models, PMs can take their problem-solving skills to the next level and drive innovative solutions for their products.

Hatch New Ideas with Glasp AI 🐣

Glasp AI allows you to hatch new ideas based on your curated content. Let's curate and create with Glasp AI :)