Data Product Manager: Do You Need One, Or Need To Be One?
We are living in the golden age of data, so it came as a surprise to no one when the discipline of data product management emerged—a PM role specializing in managing data as a product. Like other PM specializations, this job title sounds pretty compelling. But what is it really about?
So before you start applying for every data product manager job on LinkedIn, let's jam on the intricacies of this relatively-new profession and see if it's worth changing the course of your product management career path.
What Does A Data Product Manager Do?
Data product managers rely heavily on data to develop their products. They also make sure that everyone in the leadership and product team is aware of the most recent KPIs and data-based learnings to help them make better strategic and tactical decisions.
Another important role that data product managers occupy is closing the gap between data scientists and everyone else in the company (like marketing specialists, software engineers, product designers, etc.) who consume the findings that the data team produces.
Data Product Manager Roles & Responsibilities
The typical list of responsibilities for this position looks very similar to what a regular product manager might have. However, everything they do heavily relies on data.
Let me go over a couple of these responsibilities to illustrate this notion of being “data-first."
Create and maintain a product backlog
Data PMs use the findings from data analyses to identify potential new features, plan improvements to existing ones, and define the way they should look and behave.
Example: Translating Insights into Backlog Items
If you were a data product manager working on LinkedIn’s mobile app, you might look at the performance of different notifications on the LinkedIn platform. Imagine you notice that the open rate for the notification reminding your users about unread messages is peaking during lunch hours. This will give you an idea (and a feature on your backlog) to show these notifications during these peak lunch hours.
Define and constantly refine user personas
Data product managers will work with their analytics team to find common characteristics and behaviors among the users of their product and use these insights to create data-informed cohorts and user personas.
Example: Using User Behavior to Inform Personas
Imagine that you were working at Spotify as a data product manager and started analyzing the way different users listen to music. When collaborating with your data science team to group users by their behavior, you might find out that there are three distinct groups:
- Those who listen to energetic music in the mornings (probably joggers).
- Those who listen to chill music from 8 AM to 5 PM during weekdays (likely folks who are working at a desk).
- And those who prefer electronic and dance music on Friday and Saturday nights (most likely party organizers or participants).
Voilà! You have just found three potential user personas for your product, and you can go ahead and document their specificities and build features for them.
Set OKRs and KPIs
Data product managers will go above and beyond here and, instead of simply setting a couple of basic metrics to track and goals to achieve in the next quarter (like "increase retention rate by 5%,"), they will build an entire framework for tracking and acting upon metrics.
Two prominent ways of creating these frameworks are by either setting a North Star metric for your product or building KPI trees.
Example: The KPI Tree
A KPI tree is a visual hierarchy between different metrics that you want to track for your product. You usually start with the main metric that you want to improve, then define several sub-metrics—the growth of which will result in the growth of your main metric.
You repeat this exercise a couple of times until you get a tree that clearly shows the relationships and dependencies between your KPIs.
Here’s what such a KPI tree looks like for the money transfer service Wise:
As we can see, the main purpose of breaking down your main metric into KPI trees is to discover the low-level metrics that are fairly easy for you to improve (e.g. fixing the conversion rates of sending a transfer).
To sum up, data product managers are the evangelists of data-driven decision-making in the company.
Now that we've covered the nature of this profession, let’s go ahead and answer the question you might by dying to ask—"Do I have the necessary skills to become one?"
What Are The Skills Of Great Data Product Managers?
As the name implies, data product management combines data science and product management. Therefore, we can naturally assume that a great product manager will have a skill set representing both disciplines.
It won't take long to validate that assumption if you start looking at the job listings for data product managers. Here’s one that an American IT Company Blend360 has posted via SmartRecruiters.
Based on my own experience, here are some of the most important skills for this position and why they're worth polishing up before you start sending out resumes.
Data Analysis
Most data product managers work with a team of data analysts. In this case, knowledge and skills in analyzing and reading data are a major asset for better communicating and working with them.
However, there is also a chance that you are the only person in charge of data analysis in your company (this is common in small startups). In this case, it would be you who would directly work with, manage, and analyze the company data—making data analysis skills a must for this position.
Product Development
Naturally, you won't go very far as a product manager if you have no idea how to lead a product development team to build something that will inspire awe in your users. Some important sub-skills here include user empathy, exceptional communication, prioritization, strategic thinking, and experience in product lifecycle management, SaaS pricing development, A/B testing, and the other standard functions related to the product development process.
Critical and Analytical Thinking
As part of your job, you will constantly be looking at a bunch of noisy data and filtering useful information out of it. Therefore, you will need to have a strong analytical mind capable of seeing patterns and relationships in raw data.
Moreover, you should ideally be able to think critically and evaluate the information and findings in front of you with objective criteria.
SQL/Python/R
Finally, you will most likely directly interact with the data sets of your product in order to analyze its structure and offer improvements, manage and clean the data stored there, make ad-hoc analyses, and for many other reasons.
Therefore, it's important for you to know how to write and run SQL queries to access and manage the data at your disposal. For more complex tasks involving data processing and various calculations, SQL might not be enough, and you might need to write scripts using a generic or specialized programming language.
In the case of working with data, the most popular generic language is Python and the most popular specialized one is R.
As we can see, the nature of data product management dictates that the professionals in the field need to have both product-oriented skills, like a sense of user empathy, and data-oriented skills, like fluency in SQL and Python.
At this point, I assume you might have a pretty firm grasp of what data product management is about.
But I also expect you to argue that data-driven activities are important for ordinary product managers, too, and wonder what the difference is between a regular PM who makes data-driven decisions and a specialized data product manager.
Data Product Manager vs. Product Manager
The difference between these two roles in the company can look fuzzy at first. However, if you start looking closely at some of the key aspects of these professions, you'll notice some significant differences.
Let me go over a couple of these aspects and show you how the work of traditional and data product managers differs.
The Areas of Responsibility and the Goals They Pursue
Similarities
For both roles, the main goal is the development, growth, and overall business success of the product. In both cases, the user is at the center of their attention, and all of the features and improvements are aimed at solving user pain points.
Differences
If you remember the famous Venn diagram for product managers' skills and focus areas, here is what it would look like for both professions.
Here, we can see that in both cases, the technical knowledge is important, as both types of product managers will eventually work with software engineers to make their vision a reality.
The same rule applies to business knowledge, too, as the efforts of both should lead to business impact. The third point, however, is where we see the main difference.
Data product managers will pay special attention to the type of data you gather from your users. They will also constantly assess the quality of the data and suggest improvements to the current database architecture and the user-facing product features that gather this data in order to keep the quality at a high level.
Traditional product managers, on the other hand, will make sure that the user journeys are well-developed and optimized. They pay attention to the way users interact with their features and can easily navigate through the application.
Reliance on Data
Similarities
Yes, I agree—product managers of any kind should pay attention to data and ensure that their decisions are data-driven. Both types of product managers will work on setting up a data analytics tool for their product and constantly monitor key product metrics such as the funnels for activation and conversion, retention, feature discoverability, and more.
Differences
For traditional product managers, data is one of the sources of learning and information they use in their day-to-day activities. Apart from relying on analytics and SQL queries from the product database, they are also obtaining information from other sources, such as customer surveys, user interviews, usability testing sessions, and others.
Data product managers, on the other hand, live and breathe data. Databases and the analytics tools and dashboards built around them are the principal source of learnings and customer insights for them. While traditional product managers are data-aware, data product managers are data-driven and data-first.
If you consider yourself pretty data-obsessed, becoming a data product manager undoubtedly sounds like a tempting prospect. But is it a job that pays the bills? I'm so glad you asked.
What Does the Job Market Look Like For Data Product Managers?
Data product managers are highly valuable professionals who can have a great deal of impact on the growth and success of a software business, and their importance is directly reflected in their salary range.
According to Glassdoor, the median income for this position in the United States is around $130,000.
It includes a $103,000 base pay along with around $27,000 in bonuses and other benefits. This number can vary greatly depending on the industry and the experience of the data product manager.
If you are a rock star who can bring significant knowledge and experience to the company hiring you, you can ask for up to $200,000 in compensation for your work.
These numbers also differ based on the country where the company seeking a data product manager is situated. In the British job market, for instance, the median salary is around £63,000 or ~$78,000 with the current exchange rate.
As we can see, the salary range that companies offer for the position of data product manager is quite high. But what about the prospects of this profession in general? Do data PMs have a bright future?
Uh, yeah. They do.
We live in the age of data, and many macroeconomists consider big data to be the new oil.
There is a ton of growth and business potential hidden in the millions of rows of data stored in your product’s databases.
There is also the recent gold rush with respect to building and using machine learning models for various tasks in products. A data product manager can be of great help here, too, as they own the quality and usability aspects of the data and can help AI engineers build better machine learning algorithms.
Company’s Point of View: Why You Should Consider Hiring A Data Product Manager
All of this is great news for aspiring and current data PMs—but what about employers? Based on the salary estimates, it seems like the data product manager role can be costly to a company's bottom line, especially for smaller startups. So, you might ask if it's really worth hiring a data product manager in the first place.
Before I begin listing the reasons why hiring a good data product manager is a rock-solid investment, I need to do a quick disclaimer.
Data product managers are valuable in companies with a lot of data. If you are still a tiny startup that has just started onboarding its first customers, or if your product does not generate much data by nature, then you might be better off hiring only traditional product managers.
However, if you do have considerable data at your disposal, then data PMs will help you:
- Better discover customer needs and improve user experience by analyzing their behavior and finding common patterns.
- Improve the effectiveness of your data engineers' work by communicating the business and user needs and priorities with them and guiding them in the right direction.
- Leverage data when defining your product strategy by putting lots of valuable findings on the table when the leadership team and stakeholders make strategic decisions.
- Help you increase your product metrics by identifying areas for optimization and developing different initiatives to get it done.
In short, data product managers will help you get the most out of your data.
PM’s Point of View: Is it Worth Transitioning To Data Product Management?
"So," I hear you asking, "If this job is so well-paid and valuable for software companies, should I consider pivoting my PM career and specializing in data product management?"
It depends. The work that data product managers do is both interesting and impactful. However, it is not something that everyone will enjoy—or, frankly, be any good at.
Before deciding to dip your toe into data product management, you should take into consideration the following:
Do you enjoy the customer-facing activities that product managers do?
I’m talking about conducting user interviews, setting up and running Discord communities, and other soft-skill-heavy functions. If you enjoy these things, consider that data product managers rarely engage in them. They learn about their customers by analyzing data instead of talking to them directly.
Do you enjoy spending hours on problem-solving tasks?
If yes, then data product management is a great choice for you as you will constantly face different kinds of logical problems that you need to analyze and solve.
Are you comfortable working with query and scripting languages?
If you are not comfortable writing code, you'll have no trouble as a traditional product manager. Knowing Python or SQL is a “nice to have” in that world.
With data product management, however, you will have no choice but to write queries and scripts almost every day. If that sounds like a dream, you might have found your calling!
Data is arguably the most valuable resource on the planet now.
Data product managers can translate business needs into SQL queries and come up with valuable ideas for improving your product and delighting your customers. So maybe you need one—or maybe you need to be one.
For more insights on all things product and product management, don't forget to subscribe to The Product Manager newsletter.