Product Manager Archetypes

There are four Product Manager types in modern software development. Yet, most job descriptions and career ladders drop all PMs into one bucket. Software organizations differentiate between front-end, back-end, mobile, machine learning, security, embedded, DevOps, and other engineers. It’s time for organizations to become more sophisticated about product management and do the same.

In my experience, there are four archetypes of Product Managers: Feature (Business) PMs, Growth PMs, Technical PMs, and Data/Machine Learning PMs. Over the course of their careers, some PMs evolve beyond their initial archetype and can comfortably operate in more than one type of role, similar to a full-stack engineer.

I like to classify PMs based on where their superpowers are most likely to be. Feature (Business) PMs frequently come from backgrounds in Marketing, Consulting, Design, or User-Experience (Human-Computer Interaction). This is also the PM archetype people pivoting from sales, marketing, and customer success teams will likely succeed in. Their superpowers tend to revolve around Business, Business Strategy, and User-Experience.

Growth PMs frequently come from performance marketing (e.g., conversion rate optimization, paid acquisition, or life-cycle marketing) or from data (e.g., data analyst, data scientist). They understand the intricacies of growth loops and can dissect any step of the customer journey from the data and user-experience perspectives. As such, their superpowers tend to include growth skills and user experience.

Software engineers tend to find the transition to a technical Product Manager the easiest. Their software engineering and data backgrounds give them a decisive advantage in the technical PM role. In companies where the target person is a developer, frequently, the line between a TPM and a Feature PM can be blurry, and PMs tend to wear both hats at the same time.

Data and Machine Learning PMs are the latest addition to the Product Management field. First data PMs emerged in advertising, e-commerce, content, and two-sided platforms. Bidders, recommendation systems, and matching algorithms required PMs with an understanding of the shallow models and a good understanding of a domain to identify data sources that powered those models. Most PMs come from technical backgrounds such as mathematics, physics, data science, or software engineering with a machine learning specialization. Their two superpowers include data science and business strategy. The latter is critical to help identify the optimal data sources for solving the problem and understand where AI/ML can be applied to enhance the product.

What all of the PM groups have in common is the ability to operate within and help the teams advance their missions using Discovery (e.g., Continuous Discovery) and Delivery (e.g., Agile) processes.

Below is an overview of the four core superpowers.

Growth Superpower:

• You have a good understanding of marketing analytics and proficiency in setting up customer tracking and analyzing the tracking data.
• You understand growth loops (marketing funnels) and customer lifecycle optimization well. You can identify problems in the funnel and prioritize solutions leveraging product as well as other channels (in-app, SMS, email, push notifications, retargeting, high-touch human support.
• You have a good understanding of customer mental models and the ability to identify how different elements of the user experience affect the customers’ motivation.
• You can identify friction in customer flows, simplify UI and remove unnecessary steps in customer journeys.
• You understand the core principles behind habit formation and experience implementing strategies to increase motivation and reward customers for their behaviors.

Data Analysis and Data Science Superpower:

• You can perform data analyses independently. You know SQL and Python (Pandas, NumPy, MatplotLib, SciKit-Learn). You can do data aggregations and run basic statistical analyses (mean, median, distribution, etc.). You can work with time-series data and conduct cohort analyses.
• You can design A/B tests and analyze results. You are familiar with frequentist A/B tests. Familiarity with Bayesian A/B testing is a plus.
• You understand shallow machine learning models (regression, random forests, SVM, Nearest Neighbor, K-means) and have a good sense for matching a problem to a specific model.
• You understand the machine learning model lifecycle from research to production and can guide the team across the process.

User Experience Superpower

• You are comfortable conducting qualitative (e.g., interviews, usability tests) and quantitative (e.g., research surveys) user-research studies without supervision. You are comfortable identifying research goals, including potential assumptions (hypotheses) that must be verified. You can define study targets and recruit participants. You are comfortable with different interview types (e.g., Switch Interviews), anthropological studies to understand customer context and workflow, product usability tests, and validation interviews.
• You are comfortable using product analytics to understand product performance and customer satisfaction with different features and to identify potential UX problems (learnability, memorability, usability, error prevention, and error recovery).
• You understand the principles of Human-Computer Interaction (how perception and memory work). You are familiar with key product design principles such as discoverability, constraint, feedback, visibility, mapping, affordance, and consistency. You can comfortably discuss their use with a product designer to achieve key product goals (e.g., using discoverability to increase the adoption of a feature or improving mapping to increase the learnability of a product).
• You are familiar with key psychological concepts that drive product adoption and usage, incl. the design of habit loops, BJ Fogg’s model of Behavioral Change, Self Determination Theory, Flow, Intrinsic and Extrinsic Motivation, etc.

Technology Superpower:

• You have a working understanding of web architecture. You understand core design and architectural patterns.
• You are comfortable reading APIs and discussing API design.
• You have a basic understanding of databases, caching (e.g., Redis), and messaging systems (e.g., Kafka) and understand the high-level cost implications of different design choices.
• You are comfortable reading code and discussing business logic in pseudo-code.
• You have a good understanding of modern DevOps principles and architectures, and you are familiar with key tools used in development (e.g., CI/CD, monitoring, login, containers, VMs, etc.)

Business Strategy Superpower:

• You are comfortable designing a value proposition (e.g., using the value proposition canvas) behind the product and building a communication strategy around the value proposition.
• You understand the key monetization models and can match pricing models to align value with customers. You understand customer and unit economics and how they can affect breakage when pricing is not associated with a usage/value metric.
• You are familiar with different business models and elements of a business model (Business Model Canvas).
• You understand how moats can be built around business (network effects, data network effects, brand, and building switching costs with embedding, personal data, and habits).
• You can estimate market sizes for new products and features, incl. Total Addressable Market, Serviceable Addressable Market, and Serviceable Obtainable Market. You are comfortable with top-down and bottom-up sizing.