What is Product Analytics?

Product Analytics is a tool specialized for product people and their needs. It is a smaller part of an organization’s greater BI stack (and Insight Infrastructure) that automates analyses that are frequently performed for product management.

This class of tools helps you understand “who” did “what, when, and where.” It complements qualitative methods and tools that helps you understand “why and how” a user did something.

Product Analytics can help a product manager, engineer, designer, analyst, marketer, or executive understand things like:

• What aspects of my product are users engaging with?
• What are the opportunities to improve the experience?
• What are the opportunities to grow the business?
• What are the patterns of user success or failure?

Types of analyses typically offered:

• Growth analyses that help you visualize sources of user growth (e.g., breakdown of your weekly user base by new, current, dormant vs. churned users), including where users came from (i.e., acquisition sources)
• Aggregate engagement stats show insights into active users, and frequency and duration of sessions (e.g., DAU, WAU, MAU, stickiness, session length)
• User segmentation to understand who your users are and what they have in common (e.g., percentage of users in a geography, using a certain platform or device, demographics such as gender, age, status)
• Event segmentation shows aggregate stats on user actions (e.g., number of registrations, subscription purchases, messages sent)
• Individual user details such as a clickstream (see chronological order of actions that a user took) and CRM-like capabilities show details on the user (e.g., demographics, status/state, preferences, cumulative actions)
• Funnel analysis shows you data on the passthrough (and dropoff) and conversion of users who have gone through a funnel you define (e.g., signup, onboarding, activation, purchase)
• A flow analysis allows you to see where a user goes/does after a starting point (e.g., next five actions that a user did after visiting landing page) or what the user did before an ending action (e.g., previous five actions user did before buying in-app purchase)
• Retention (and churn) analyses is critical to product growth and various methodologies can be applied: n-period, unbound, bracketed periods (e.g., what percentage of users come back to my product next week?). Retention is a lagging indicator of habituation and (typically) a leading indicator of desired results such as customer satisfaction, revenue, or rate of word of mouth marketing
• Financial metrics including customer acquisition cost (CAC) per channel or campaign and customer lifetime value (CLV)
• Other product-specific intelligence, including predictive analytics, helps you discover noteworthy patterns such as what distinguishes users that churn

Product Analytics tools are designed around a user who generates n events. A user has properties (like a CRM) that are automatically captured by the tool (e.g., location based on IP address, device used) or manually defined by you (e.g., user’s age, type of user).

Events also have properties. They aren’t used as often as user properties but can capture valuable metadata about the event itself (e.g., message sent from driver to rider is the first time they are messaging) or state of the user when the event was generated (e.g., Jane was 21 years old when she registered a year ago but is now 22).

You can think of analytics as having three layers:

• Data: collecting, processing, and management of raw data
• Analysis: allows product person to define parameters of analysis to generate report (e.g., which event done by which users during what time frame)
• Reporting: allows product person to see the report as a chart or table, configuring how the data is visualized (e.g., line graph of home page visits by unique users)

Most end users on your team will only experience data visualization (i.e., Analysis and Reporting layers).

See how product analytics fits into a larger Insights Infrastructure that you should build.