The Role of the Digital Product Manager

Ask ten Product Managers what they do, and you’ll get ten different answers. It’s time to fix that.


Leaders offer clarity: clarity of vision, clarity of role, clarity of execution. If product managers are leaders within a company, why are we so unclear as to our own purpose?

The chameleons of product development

You’re “the mini CEO.” Somebody who “influences without authority.” The “champion of the customer.”

We’ve all heard these descriptions of what we do, but beyond oft-repeated slogans there’s a frustrating lack of agreement on the specifics. Wikipedia defined product management as:

an organizational lifecycle function within a company dealing with the planning, forecasting, and production, or marketing of a product or products at all stages of the product lifecycle

Product managers “deal with” planning, forecasting, production, and/or marketing…but that covers a wide range of functions. And what exactly are PMs supposed to do as we “deal with” these things?

The role may consist of product development and product marketing, which are different (yet complementary) efforts…[the role] spans many activities from strategic to tactical and varies based on the organizational structure of the company

Oy (emphasis mine). Product management is, indeed, a role that touches many parts of the product development process. But it’s clear that the role isn’t consistent from company to company, a fact that many people agree with.

Let’s draw from our PM bag of tricks and survey the market. Here’s what some of the biggest authorities in the field have said about the role:

[to] ensure a consistent stream of innovative products and services that customers love, and that deliver on the business strategy

Silicon Valley Product Group

to discover a product that is valuable, usable, and feasible

— Marty Cagan, Inspired: How To Create Products Customers Love (the closest thing we PMs have to a Bible)

the intersection between business, technology and user experience

— Martin Eriksson, Mind the Product

[to guide] products throughout the execution cycle, focusing specifically on analyzing, positioning, packaging, promoting and tailoring our solutions to all the markets where [your company] does business

Google Careers

[to be] responsible for the management of new or existing products within an organization or business

— Kasia Mikoluk, Udemy

to make sure that the product not only matches up to customer expectations, but are (sic) also in line with the overall strategy set forth by the company’s founders and CEO

— Ken Yeung, TNW, reporting on an event organized by Google Ventures’ Ken Norton and Greylock Partners’ Josh Elman

And finally, “Product Management is Everything.”

Moving toward clarity

We “discover,” “rest at the intersection,” “guide,” “manage,” and “align.” Of the sources I polled (and these were only the first seven) that’s an ironic lack of alignment. Alas, when there’s insufficient alignment after substantial debate, it’s our job to enforce clarity and common purpose.

Product management can touch virtually the entire product development stack, from market discovery, to marketing, to sales, to release planning, to project management, to design, to engineering, to quality assurance, and finally, to deployment.

But that’s what we can touch (I’ve touched all of the above at one time or another, even telling dev ops to roll back a release before). In doing all of these things, sometimes we’re performing the role of Product Management, and sometimes we’re overlapping with others or filling in gaps.

Rather than saying what we can do, I define here the kernel of product management that every single PM organization should perform no matter the industry or culture.

The kernel of product management

The Product Management organization has a single responsibility:

to maximize the company’s long-term market value through the research and development of new products

If that sounds far-reaching, that’s because it is: there’s a reason we’re sometimes called mini-CEOs. Indeed, often the actual CEO is the ultimate product leader (e.g., Zuckerberg, Page post-Schmidt, Gates pre-Ballmer, etc.). If a company isn’t realizing its potential in terms of market ownership — for whatever reason — the leader of the product organization’s neck is on the line. [Note for the pedantic: “new products” include iterations on existing products and even, sometimes, the removal of features. If it’s materially different, it’s “new” for our purposes.]

In fulfillment of that responsibility, we have a single mission: to deeply understand our customers’ needs and ensure that they are uniquely and sustainably addressed by the company.

Finally, in service of that mission we perform what I call “the five P’s”. We:

  • Paint the product vision
  • Plan product releases
  • Push for alignment, quality, and clarity
  • Pull the trigger
  • Performance track

We can also (and usually do) perform project management, but it’s not an essential part of the job (some companies sequester this function into the Engineering Project Management, or EPM, role). We can also design product features and releases, but again, that is not an essential part of the job (this function is increasingly being sequestered in UX, UI, visual design, and other roles within a Design organization that often but does not always report to Product).

This definition is broader than the one outlined in Inspired (by Marty Cagan, referenced above), which limits the PM’s role primarily to that of product discovery and definition (best aligned with Paint). Startups in the Bay Area have, I believe, expanded the role of product management beyond that which evolved in larger companies — as Program Managers at Microsoft, for example, we primarily wrote product specifications, with planning the responsibility of a team (of which the PM was the leader, but it was still a team). My best guess as to why this change has happened is that the startup ecosystem is rapidly embracing collaboration with a decider over consensus building as a superior mode of decision-making. I expect this change to trickle up to the Microsofts of the world over time, as new best practices from startups often do.

In subsequent essays, I’ll dive deeper into what each of these functions entails. But, briefly:

Paint the product vision entails understanding the intersection between what your customers want and what your company can produce, and then communicating that to the rest of the company (most commonly in the form of Market and Product Requirements Documents, although these are often perceived as too heavy for startups — Amazon famously uses shorter press releases). Sometimes Product is also responsible for communicating externally. Painting a coherent product vision requires understanding the design and architecture of what you’re going to build, but it does not require being able to produce them yourself.

Plan product releases entails setting a schedule for what the company is going to release and when. One artifact of this process, the Roadmap, is the deliverable most commonly associated with product managers. Planning requires gathering information from the market (directly or in partnership with Sales and/or Marketing) and from R&D (including ideation from Design, feasibility assessments from Engineering, etc.), and then ruthlessly prioritizing initiatives and activities.

Push for alignment, quality, and clarity entails running around like a chicken with its head cut off until everybody understands what they’re building, why they’re building it, and that it’s built well and on time. This involves constant triage of exceptions to the plan, since nothing goes as you expect it to.

Pull the trigger entails wielding the ultimate responsibility to decide to make a new product go live, and then ensuring successful delivery of the product to its consumers. The most straightforward of the P’s, but also usually the most frightening.

Performance tracking entails deciding upon key performance indicators for a new product and tracking those KPIs post-release. The state of the art has drastically improved in quality over the past few years, with excellent new tooling and methods standardization (e.g., cohort analysis, a/b testing, universal analytics, etc.).

One of the common mistakes that new product managers make is to strive for perfection. They worry that there’s an objectively “right” thing to build that will eventually reveal itself. Product management, however, is as much an art as a science. Painting a vision and planning a roadmap are remarkably similar to my years as a scientist in graduate school, in fact: you build a model of the world (in this case, a psychological and economic model of how people will interact with your product in the market), you create a hypothesis based on that model (we’ll see a 10% increase in installs and a 15% increase in sharing activity), and then you run an experiment. Product leaders and investors even use that terminology directly: “run the experiment.” Because no matter how good you are, you never really know what’s going to happen. The best product managers are ones that are perceived as credible across the entire company: a great PM can credibly speak about what customers want and how the company can give it to them. Move fast enough, keep iterating, and as long as you’re tacking generally in the right direction, you’ll get closer to the goal.

A final note on “influence without authority”: let’s put this misconception to rest. Product managers may not always have hiring and firing rights (although sometimes they do), but even the CEO rarely exerts this authority over non-direct reports. Deciding what gets built, and when, and approving a release to your customer base is more than enough authority to keep you up at night.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.