What’s a “Data PM”?

Introduction

Hello, world — I am Olga, a Data Product Manager within the Go-to-Market (GTM) Data Engineering team at Atlassian. The purpose of this blog is to shed some light on my role with examples of workstreams owned by similar roles in the data space. We will cover a few examples of data products and services within a larger data landscape, and how all this clicks together.

• Opinions are my own and don’t represent Atlassian as a whole

A long time ago, before “PM” roles existed,

there was a business role opening listed.

With a hope for the best and wide-open eyes,

I plunged into it to keep asking the “Why?”

Jokes and hardly-profound poems aside, some of the critical elements of a Product Manager’s job can be summarized with an ability and willingness to ask questions. Why? Because the quality of the answers we receive correlates with the quality of the questions we ask. The more curious a PM is, the better. Even some PM job descriptions these days include a nice-to-have requirement to “fearlessly challenge the status quo.” Why? Because curiosity matters.

Credit: Airfocus on LinkedIn

My wonderful colleagues at Atlassian have expressed interest in learning more about my personal and professional experiences as a PM, to which I could not say “No”. On the other hand, saying “No” to unprioritized work is one of the key strengths of a PM, however, the role itself has evolved over the last decade.

Product Manager role

So, what does a Product Manager do? To answer this, here is a list of questions I often get asked:

• Are you a project manager?
• A program manager?
• What is a product manager?
• What kinds of products do you manage?

Of course, there is no single definition of a PM role as it varies between organizations, industries, and countries; however, some patterns still apply. Here is my user-friendly explanation:

• Product ≠ project ≠ program manager.
• Product → The Why: we drive features, enhancements, value-add, and cross-functional efficacy.
• Project → The When: these fine folks are timekeepers and Gantt chart wranglers.
• Program → The How: PgMs are all about the process, the flow, and the people.

When it comes to roles & responsibilities, we rely on custom playbooks developed by Atlassian to get alignment within each team and cross-functionally:

How to define roles and responsibilities for team members

Within my team — Go-to-Market Data Engineering — the most recent discussion about “What does a Product Manager do?” led to the following results:

• Engage and collaborate with stakeholders
• Partner with the engineering manager to work out the capacity
• Own the roadmap, timelines, tradeoffs, and the feature backlog
• Evaluate and showcase the value of work delivered
• Sprint planning and grooming in Jira
• Propose product enhancements/features

This overview would not be complete, however, if we didn’t add other perspectives to the mix:

At this point, we may’ve been able to decode the role of a PM a bit. Now, what does a “Data PM” mean? Is it significantly different from a broader “PM” term? Or from a UI/Business/Technical PM?

As always, there are more questions than answers.

Data Product Manager role

Since the data domain can be chaotic due to its ever-evolving nature, data quality issues are unavoidable. Code bugs will creep up, and discrepancies between the data sources and data visualization tools will arise. There will be human errors and systematic legacy errors. Tech debt will find its way into the beautifully-orchestrated pipelines. The role of a Data PM includes all of the baseline PM attributes and expectations, with a few added layers of data-relevant responsibilities:

• Partner with business teams in setting the vision for data needs and translating it into a product roadmap and end-to-end execution
• Knowledge of data warehousing concepts, data pipeline design, experience in writing SQL or Spark
• Craft and execute on a product backlog, balancing innovation with maintenance and support
• Develop and defend strong points of view with data and analytics
• Work with structured, unstructured, and graph data

One of the most resonating quotes in my Data PM career has been this (source unknown):

“Help your team ship the right product.”

Why? Because each segment of this sentence has an impact and, based on where the focus is, it brings a different flavor. Let me unpack this.

• Help your team ship the right product: a Data PM brings different teams together as a mediator or “glue,” helping everyone reach common goals. It’s not uncommon to balance out conflicting priorities in the backlog, so strong data points supporting one decision versus the other is a Data PM’s best friend. We use a DACI framework at Atlassian to ensure alignment is reached between all parties.
• Help your team ship the right product: there is no “us” vs. “them” in product: everyone is a “team” as long as there is a bigger picture everyone can contribute to. Since data is a backbone for virtually all teams in the SaaS landscape, the Data PM is responsible to drive cohesion between Marketing, Sales, Customer Support, Legal, Finance, People Ops, Martech, Salestech, and other business partners.
• Help your team ship the right product: the keyword here is “ship,” which is often more beneficial than “achieve perfection.” Shipping is progress. Perfection is a myth. Continuous progress is the most rational way to get close to perfection. While data is not as visual as UI enhancements on a website or marketing copy improvements, incremental improvements are critical as it helps teams iterate and optimize existing campaigns and growth initiatives further.
• Help your team ship the right product: while it’s subjective, there is no “right” or “wrong” — shipping a product is all about the highest value-add with the lowest risk and cost. Simple, isn’t it? (It isn’t). Once again, incremental value-add and data-driven decisions ensure we ship more “right” than its alternative.
• Help your team ship the right product: finally, we’re back to the Product topic again! In this context, “product” is a term to describe the outcome delivered compared to the expected/preferred outcome. The goal is to ideally “underpromise, overdeliver” and ship data products and services to maximize stakeholder and customer outcomes while minimizing manual effort and costs.

Data products and services

When it comes to data, the concept of “products and services” may seem confusing and/or intimidating. Indeed, data supporting the lifeblood of operations and marketing is usually behind the scenes. At the same time, it’s all about perspective: in the SaaS world, anything can be seen as a data product and/or service. Not convinced? What about -

• Various data pipelines supporting marketing, sales, finance, people operations, data science, machine learning products, and more
• Third-party vendors reporting tools to supplement ROI analysis, ad spend, and marketing campaign effectiveness
• Ad vendor tracking and adtech
• Web data & social media integrations to deliver insights on social engagement with Atlassian products
• Web event tracking and instrumentation to measure user engagement with our marketing content, webinars, courses, public roadmaps, and various other platforms
• Data visualization tools — out-of-the-box in the data lake and custom-built — to showcase the value of work, trend analysis, patterns, and behaviors to predict and even prescript future initiatives and roadmap action items.

GTM Data Engineering covers all these data products and services working with various teams from marketing to IT to finance to legal, while other data teams focus on personalization, customer experience, creative content, product marketing, and more.

There is a close correlation between data products and projects, too. An example would be a new third-party integration, such as a data ingestion tool, and its outcomes. While the timelines of the projects including research, development, testing, quality assurance, user acceptance, and documentation would be part of the project; the outcomes such as final pipeline delivery, its usability, operability, stability, and data quality would be considered data products. Yes, close collaboration between multiple stakeholder groups and business partners is required and necessary for a successful outcome.

Piece of advice

Don’t lose heart. Believe in everyone’s best intentions. Look for opportunities even when everything is calm and quiet. Especially when everything is calm and quiet.

This philosophy of mine corresponds closely with Atlassian values. Sounds relatable? Check out open PM roles at Atlassian— don’t miss the opportunity to explore the possibilities!

#InDataWeTrust