What is data product management? How is it different from product management? Why is it important?

These are just a few questions we’re answering in the article.

Intrigued? Let’s get to it!

TL;DR

• Data product management is the discipline of collecting and analyzing data to develop and improve products.
• Data products are built around advanced data processing, AI, and machine learning.
• The data product manager role dates back to the early 2000s when companies realized they needed dedicated data professionals.
• Data product managers are responsible for collecting, analyzing, and sharing data within the organization in accessible ways to inform product strategy.
• Data product managers combine the roles of product managers and data scientists.
• You need data product managers when your product or your organization’s internal processes depend on the effective use of data.
• Apart from regular product management skills like leadership or UX design, data product managers need to be proficient in managing the data product lifecycle.
• A data product management team has two groups. One processes the data while the other uses data to develop solutions that meet user needs.
• Data PMs use data to set goals, build roadmaps, and validate ideas. They also track product usage and conduct A/B tests.
• Usage tracking tools allow data product managers to measure the value of different features. They can also use them to identify the best ways to guide users to value.
• A data PM can use UX analytics tools to record and analyze user behavior in-app to optimize the UI.
• AI and machine learning tools help data teams predict user needs and design ways to satisfy them.
• AI can also be useful to find patterns in user feedback to boost retention and improve user satisfaction.
• Monthly and daily active users (MAUs and DAUs), conversion, churn and retention rates, the NPS and CSAT scores, and the Customer Acquisition Cost are important metrics data product managers should follow.
• Interested in how Userpilot can help you collect and analyze data? Book a demo with us!

What is data product management?

Data product management is the discipline of harnessing data to develop and refine products.

It applies data science techniques and data engineering processes to the data collected during market research and product discovery to guide the development of new features or the improvement of functionality.

Data product management is particularly relevant for developing data products.

What are data products?

Data products, like any other products, are developed to satisfy user needs and deliver value.

However, what sets them apart from other products is the advanced use of data analytics, AI, machine learning, and a range of data management techniques.

Examples of data products are streaming services, which use machine learning to customize content recommendations for users.

A brief history of the data product manager role

The data product manager role started appearing in the early 2000s.

Companies that heavily relied on data to inform their product development, like Netflix or Uber, realized that they needed professionals whose focus would be on developing data solutions.

What does a data product manager do?

Product managers are in charge of developing data products that are scalable and accessible.

Their responsibilities include:

• conducting market, customer, and competitor research
• mapping user personas and customer profiles
• defining metrics to track progress, setting KPIs and OKRs
• performing tests and experiments to assess the effectiveness of solutions
• tracking product usage to identify areas for improvement
• developing data pipelines to ensure all teams have the data they need
• interpreting and presenting data in accessible ways
• developing data-based business cases to influence stakeholders
• promoting a data-driven mindset and data literacy within the organization
• developing a data-driven product strategy

Data product manager role vs product management role

The product manager and product data manager roles are similar in that their job is to develop products that delight users and provide excellent user experiences while helping them complete their jobs.

PMs, just like data PMs, use data to make their decisions. However, the data PM is way more independent in this respect. They don’t need to depend on the help of data analysts and are able to leverage the data themselves.

You could say that in a sense, data product managers are a cross between product managers and data scientists.

Do you need data product managers in your product team?

Successful product development is hard to pull off without data.

But shouldn’t every product manager be able to conduct experiments, collect data and analyze it?

Definitely, but their data science expertise may be limited especially when it comes to developing complex data products. If your product is heavily dependent on data, hiring a data product manager will be a good investment.

Of course, data scientists and analysts can deal with the more complex aspects of data management, but they lack UX design experience, technical skills, and business acumen. Data product managers combine all of these.

What’s more, dedicated data product managers can help the organization build its data-driven strategy and develop processes, to ensure that the data is used consistently and rigorously to inform decisions.

By developing robust data pipelines, a data product manager can increase access to information across the organization and break down teams and departmental silos.

Once you have such processes in place, your organization is less dependent on the individuals who set it up initially. This ensures continuity should they leave.

Skills data product managers need

A data product manager needs all the skills a regular PM must have.

To start with, they need solid business sense to ensure that the product helps realize the organization’s business strategy.

Next, they need great leadership skills. They are in charge of teams, so they need to be able to build good teams and guide them through change.

This is closely linked with communication skills and emotional intelligence. The latter is critical when explaining difficult concepts to people unfamiliar with data science.

Moreover, PMs need to be familiar with psychological and UX design principles to build products that are usable and accessible, and delight users.

Data product managers also require good technical knowledge to be able to communicate their ideas to the development teams and set realistic objectives. In the case of data PMs, this covers things like machine learning and AI.

On top of these, a data PM requires needs to have a good knowledge of the data product development cycle.

They must be experts in data preparation, assessment, and analysis. They need to be able to build and test hypotheses. Once the product is ready, they should be able to plan the product launch and its operation.

How to structure a data product management team

A good data product management team has two cores: the model core and the activation core.

The model core consists mostly of data scientists and is responsible for collecting the raw data, interpreting it, and creating predictions based on that.

The activation core is the product development team who uses the outputs from the model team to develop solutions that will deliver value to the users.

For a data team to be effective, the two cores need to work closely together. The model needs to take into account the input from the development team, and the development team requires a reliable forecast to build solutions.

How do data product managers use data?

Let’s explore how data product managers use their data expertise in practice.

Set SMART goals and realistic OKRs

To be able to set good goals for your product team, you need the right data and know how to interpret it. That’s where data PM’s knowledge of data science comes into play.

Skilled PMs are able to extract the right information to identify SMART goals and OKRs that are specific and can be realistically achieved within the allowed timeframes.

Leverage data to build data-driven product roadmaps

Data product managers are able to utilize data science to inform product roadmaps.

In practice, this means using qualitative and quantitative data for backlog grooming and prioritization of the product features that customers will find most valuable and which allow realizing organizational goals.

One of the consequences of making data-driven roadmap decisions may also be sunsetting features that are not delivering the intended results.

Fake door testing to collect data and inform the product strategy

Data product managers help their teams validate ideas before they are built to ensure they will deliver the desired results.

Prototypes are the classic way to do this in a cost-effective way. Fake door testing requires even less effort and money.

How does it work?

Let’s look at the hypothetical example below. Imagine that Asana is only considering the Goals feature. However, they’re not sure it will get any traction with their users, so they add it to the menu and create a small tooltip to attract users’ attention.

Of course, the feature doesn’t exist yet and users find it out immediately when they click on it. However, the click-through rate will allow them to decide if there’s enough user interest to justify it.

It’s also a great way to invite users to beta testing when the feature is finally ready.

Track product usage to understand product value

By tracking product usage and analyzing the data, data product managers can identify what exactly brings value to users – and what stops them from experiencing it.

Once they know which features and parts of functionality drive value, they can leverage the data to map out the shortest paths to success for different user cohorts.

Robust product usage data can also help teams pinpoint the reasons why users churn.

With Userpilot, product managers can track product usage and feature engagement across different time periods and user segments.

A/B testing to enhance product usage through personalized user experiences

In-app engagement flows are a recognized way to guide users through the journey and into adoption. However, no single flow will suit all user segments, so you need to test their effectiveness.

The data team can design and run A/B tests to do just that.

The key to successful A/B testing is targeting specific user segments. To do that, you need to use data, so here’s another way a data product manager can contribute to product development.

Use data and analysis to inform and influence stakeholder decisions

Influencing stakeholders is a big part of the PM role, and objective data can help with that.

If you can show how a product or feature can save time and money, improve adoption or reduce churn, you increase your chances of securing their approval.

The catch is that large amounts of data are often very inaccessible to untrained team members, even very senior ones.

That’s why it’s the data product manager’s job to break it down into digestible chunks and present it in a simple way. This could be through data visualizations.

How do data product managers collect data?

Choosing the right tools to gather the right data is the key to success as a data product manager.

Product analytics tools for tracking

Userpilot allows you to track and analyze data to determine the value of particular features.

The feature tagging functionality is very intuitive and simple. You can do it directly from the Chrome extension.

Once you set it up, you can easily follow user interactions with different aspects of the UI.

And the interactions don’t mean just clicks but also hovers so that you can tell when a user got attracted to a feature of the UI. The functionality also allows you to track when users fill in text fields too!

What if you want to track a combination of events? Like a number of actions that users need to complete to reach the activation stage? That’s no problem at all.

Userpilot allows you to group interactions into custom events so that you can track and analyze them as if they were one.

UX tools to understand user behavior

UX tools like Hotjar allow you to analyze how users engage with parts of the user interface.

The software gathers information on what users do in-app – their clicks, scrolls, and cursor movements.

Next, it creates a heatmap of most parts of the UI. Once you know which UI elements don’t attract enough attention, you can tweak them to obtain better conversion.

Hotjar also enables you to create live recordings of users’ interactions with your product.

Thanks to that, you can identify the issues that they may come across, like misleading messaging, and ways to improve them to smoothen their user journey.

Predictive analytics tools

Modern data products rely on AI and machine learning algorithms to predict user behavior.

As a result, based on the user data, you are able to anticipate their needs and design solutions that meet those needs. When the user finally experiences the need, the solution is ready and available at the exact moment and place.

AI tools like Pyoneer can be used to analyze user feedback, both quantitative and qualitative, to extract patterns. Such insights allow data product managers to respond to complaints effectively to reduce churn.

Important metrics for data product managers

What metrics should data product managers pay attention to? Some of the most important ones include:

• Monthly and daily active users (MAU and DAU): the increase in MAUs and DAUs is naturally a positive trend. The key to the effective use of the metrics is consistency. The data product manager should decide what makes an active user – what actions and what frequency.
• Conversion Rates: how many of the users that book a demo or start a free trial become paying customers.
• Customer churn rate: how many users drop out. A high churn rate may mean that the product doesn’t meet the needs of the users or they don’t experience its value in time.
• Retention rate: the opposite of churn rate, shows how many customers stay and keep using the product and pay their subscriptions.
• Net Promoter Score: the overall user sentiment of the users towards the product as a whole.
• Customer Satisfaction Score (CSAT): similar to NPS but instead of a general feeling, it focuses on a specific area like onboarding or support.
• Customer Acquisition Cost: how much it costs to get a new customer. The lower the cost, the better. The acquisition costs tend to be higher than the cost of retaining customers.

Conclusion

Data product management improves the utilization of data to develop products that delight users. It is particularly relevant for data products that depend on quality data processing to deliver value.

If you would like to learn how Userpilot can help you collect and analyze user data, book the demo!