Who This Helps
This is for junior analysts who want their work to actually get used. You run the numbers, build the dashboards, but somehow the team still makes decisions on gut feel. Sound familiar?
In Product Metrics Basics, you learn to define metrics that everyone trusts. No more "where did that number come from?" debates.
Mini Case
Meet Priya. She's a junior analyst at a SaaS company. The team keeps arguing about whether new users are sticking around. Priya runs a retention report and finds that only 12% of users come back after 7 days. But the product manager says "our retention is fine." Why the gap? Because they're using different definitions of "retention." Priya needs one clear metric everyone agrees on.
Do This Now (5 Steps)
- Pick one action that matters. In Product Metrics Basics, the first mission is "Activation Definition." You define activation as one event plus one time window. For Priya, that's "complete onboarding in 3 days."
- Write down the exact event name. Use the "Event Taxonomy" mission. Choose 5 key events and list their required properties. Example: "signupcomplete" with properties "plantype" and "referralsource."
- Set a North Star and two guardrails. The "North Star & Guardrails" mission helps you pick a metric that guides the team. Priya's North Star is "weekly active users." Guardrails: "support ticket volume" and "page load time."
- Look at one segment. The "Segment Snapshot" mission shows where activation breaks. Priya segments by "referral source" and finds that users from ads activate at 8% vs. 20% from organic search.
- Share your findings in a one-pager. List the metric, the segment, and one clear recommendation. Example: "Shift ad spend to organic channels to improve activation by 12%."
Avoid These Traps
- Defining metrics differently each week. Stick to your event taxonomy. If you change definitions, the team stops trusting the numbers.
- Showing too many metrics at once. Pick 3-5 that tell the story. Priya only shares activation rate, retention rate, and North Star.
- Forgetting to include a recommendation. Analysis without action gets ignored. Always end with "so we should do X."
- Using jargon like "cohort analysis" without explaining. Say "group of users who signed up in the same week" instead.
Your Win by Friday
By Friday, you'll have one clear metric definition that your team agrees on. You'll know exactly how to measure activation and retention. And you'll have a simple one-pager with a recommendation that gets approved. That's a win.
And hey, you might even get a high-five from your PM for finally ending the "what does retention even mean?" debate.