The Role of Product Management in Data Science

PinIt

Data science should now be part of every product manager’s general education, not so they can get into the details of “how,” but rather, so they can understand “what could be.”

Artificial intelligence (AI), machine learning (ML), and data science have become integral parts of much of the technology we use daily.

We expect our products to be smart and responsive, to know things about us, and to anticipate our needs. The popularity of these capabilities has companies scrambling to figure out how to integrate them into their product roadmaps.

See also: 5 Keys to Better Customer Experiences and Revenue

As was the case with Agile, continuous development, and microservice architectures, we now see AI, ML, and data science as en vogue, and for good reason: there is a ton of potential value in harnessing the power of data science to solve critical business challenges. Algorithms can be used to make sense of a massive amount of data, and both machine learning and AI can automate tasks that humans find tedious.

To realize these benefits, engineering teams are starting to add data scientists to their teams. Data scientists are able to build data models suited to the organizations’ needs, have an understanding of SQL, and are adept at translating data into insights.

So if engineering teams are staffing data science teams, how does the product organization adapt?

Do product managers need to specialize in data science?

Understanding — and driving — business requirements is an incredibly important function of the product team. It is imperative the product team understands the questions that need answering and shares that information with the engineering team.

When a data scientist is in the mix, they may uncover new applications based on what they’ve learned from the data. Even in that case, however, it is still on the product team to figure out how to turn that information into a deliverable.

This raises an important question: do we need product managers to specialize in data science?

There are some good arguments for a data scientist in the product manager position. Both product managers and data scientists use data to make decisions and specific metrics to measure the outcome of those decisions. A product manager needs to know what success looks like for a product or a feature, while a data scientist chooses evaluation metrics that define the outcome of an experiment. Both product managers and data scientists then need to be able to explain their decisions to stakeholders on other teams clearly. They need to be technical, business-oriented, and creative enough to communicate with everyone from engineers to designers.

Of course, whether you want a data product manager on your team or not will vary based on your industry, product, customer, and use case.

However, when I evaluated this question for my own team, the answer was “no.” I don’t want my product managers tied to only data science projects.

Business needs versus hard data

My reason for not tying product managers to data science is simple: I want the product manager focused on the business need. Product managers shouldn’t be primarily focusing on data. Instead, they should be completely tapped into business stakeholders and able to understand and articulate their needs.

I want product managers asking the following questions:

  • What will solve the customer’s problems?
  • What are our delivery challenges?
  • Where do we have blind spots that could help us operate better?
  • How can we provide greater insights to our customers?

Once the product manager identifies a problem, we can work with engineering and their data scientists to investigate options together and determine an implementation approach.

Let’s say the business challenge is the amount of time being spent on a very manual selection process. We might understand fundamentally that we want to automate the process and even be able to work with the business to identify the attributes of our search. But what data is available, and how can we use that intelligently to make selections? This is where our partners in data science can help and come up with the algorithms needed to automate.

The relationship between PMs and data

Does this mean product managers should be completely innocent of data science? Of course not.

I firmly believe that an understanding of data science should now be part of every product manager’s general education, not so they can get into the details of “how,” but rather, so they can understand “what could be.”

This understanding gives product managers a sense of the kinds of questions data science can answer, so they begin to think more creatively about solutions that would benefit their biggest stakeholders: the customers.

Kristin Simonini

About Kristin Simonini

Kristin Simonini is VP of Product at Applause. A 20-year veteran of the product management space, Kristin leads Applause’s product organization. Her team is responsible for defining the strategic roadmap for Applause's industry-leading crowdsourced testing platform. Prior to Applause, Kristin led the product management efforts at EdAssist, a Bright Horizons Solution at Work where she instituted a product management practice and led the effort to reinvigorate their industry leading tuition assistance platform including the release of their first mobile app.

Leave a Reply