Product Management Unpacked

According to a Brookings Institution report, “Automation and Artificial Intelligence: How machines are affecting people and places,” roughly 25 percent of U.S. jobs are at a high risk of automation. Among the most vulnerable jobs are those with routine physical and cognitive tasks such as office administration, production, transportation and food preparation.

The jobs that are the least vulnerable to automation are generally classified as abstract and manual occupations — “those that involve tasks that are … difficult to codify or take place in physical environments that are difficult to control.” According to the report, “abstract roles — typically in management, technology or finance — tend to require more formal education and skills such as creativity, persuasion, intuition and problem solving.”

The report predicts what automation does not replace, it will complement — as will be the case with many technology workers. Those whose jobs are supplemented by AI are likely to see benefits including increased productivity and overall job satisfaction. This prediction is already in line with how AI is influencing the role of a product manager.

In product management, AI is increasingly leveraged in the research phase of the product lifecycle. A recent Forbes article explains how AI is enabling product teams by offering unimaginable consumer insights that inform important product decisions. By analyzing internal customer service data and external review data, AI enables product managers to test their hypotheses and gauge previously unknown scenarios. The assistance of AI in product research allows product teams to focus their energy on development and also helps reduce some of the risks associated with product decisions.

While AI and machine learning can provide valuable insights regarding customer preferences, this technology is meant to enhance — not replace — the role of a product manager. Ty Magnin, Director of Marketing at UiPath shares his take in his recent Appcues blog: “Machine learning allows your app to track user history, but that doesn’t mean you can say with certainty what your users are going to do.” Instead, Magnin suggests that product managers use machine learning predictions as options they can provide to their users to choose from. “Based on what [users] choose, make adjustments to your product so that your users can see how you’ve solved a problem better than any of your competitors.”

Given the vast potential of AI with informing product development, learning AI models and tools for data analysis is becoming a valued skill in product management.

However, AI is not only evolving the role of a product manager. Increasingly, it is also shaping product features and capabilities. According to a ZDNET article, “as artificial intelligence (AI) and machine learning (ML) capabilities are designed into new products and services, product managers need to enhance their skills in order to develop and provide functional requirements and AI-powered specifications to data engineering and data science teams.”

Carnegie Mellon University’s MS in Product Management program includes courses such as Data Science for Product Managers. Students in this course are introduced to a variety of data science techniques that help inform critical product management decisions. These techniques include preference modeling, time series forecasting, regression, clustering, classification, A/B testing and analytics for unstructured data. Along the way, students learn about practical aspects of applying data science to product management, such as choosing appropriate metrics for product success.