Topic 9 — Product metrics and analytics

Goal: Learn the numbers a PM uses to know whether the product is working.

Lesson 9.1 — Why a PM reaches for a number

In a standup at Lumi, the personal budgeting app where Maya works, someone says the new weekly-budget feature is "going great." Sam, the senior engineer, doesn't look up. "Going great how?"

Nobody has an answer. They have a feeling.

That gap is the whole reason metrics exist. A metric is just a number that tells you something about how the product is doing, and a PM uses them to do three things: spot a problem early, see whether a change actually helped, and back up an argument about what to do next. "Going great" can't do any of that. "32% of people who opened the feature came back to it the next week" can.

You do not need math beyond arithmetic and percentages to start. That surprises a lot of career-changers. The real skill is choosing the right number to look at and reading it honestly. Heavy statistics is someone else's job. Picking the number that answers the question in front of you is yours.

A PM who can't read the product's numbers is flying blind, and everyone in the room can tell.

Lesson 9.2 — The number that flatters you

Maya pulls up the dashboard and sees it: total sign-ups, all time. Big, bold, climbing. It feels good to look at.

It is also useless, and learning why is one of the most useful things a PM can learn.

That figure is a vanity metric — a number that looks impressive but doesn't reflect real value or tell you what to do. Total sign-ups can only ever go up. Even if every person who joined this week quit the next morning, the number keeps rising and keeps looking like a win. It flatters you while the product quietly leaks.

Two habits keep Maya honest.

First, she prefers actionable metrics — ones that, if they move, tell her to actually do something. "Weekly active users" or "checkout conversion rate" point at a decision. "Total registered users since launch" just sits there looking nice.

Second, she pairs her numbers. One figure alone can lie. Lumi once saw "messages sent" jump and nearly celebrated, until Sam pointed out a bug was forcing people to retry failed sends. The number went up because the product was broken. So she pairs a metric with a counter-metric, like watching engagement and complaints together, to catch that kind of trap.

Before you trust a number, ask: what would this make me do if it moved? If the answer is "nothing," it's vanity.

Lesson 9.3 — A map of where users fall off

Lots of people sign up for Lumi. Almost nobody is still around two weeks later. Where, exactly, is the leak?

You need a map of the journey to answer that, and the most popular one has a goofy name: the AARRR funnel, also called pirate metrics (say the letters out loud). It breaks the customer journey into five stages, in order:

• Acquisition — how people find and arrive at your product.
• Activation — whether new users reach their first real value, the "aha" moment.
• Retention — whether they keep coming back. Often the single most important stage.
• Referral — whether they tell others.
• Revenue — whether they pay.

The power of the funnel is that it shows you where you're winning or losing people. One overall "good" or "bad" number can't point at a stage; the funnel can.

Take Lumi's problem. Tons of sign-ups, nobody returning. That is not an acquisition problem. Acquisition is the one stage that's working. The leak is in retention, and pouring more people into the top of a leaky funnel just wastes them faster. Pinpointing the weak stage tells Maya where to spend her energy, which beats staring at one big overall number that hides the whole story.

Lesson 9.4 — The vocabulary you'll hear in every meeting

In Maya's first month, Dan rattled off "our churn's creeping up but the LTV-to-CAC still looks fine" and moved on, and she nodded like she understood. She didn't. These words get thrown around constantly in product meetings, so it's worth pinning down what each one means and what question it answers.

• DAU / MAU — Daily and Monthly Active Users. How many real people use the product in a day or a month.
• Churn — the percentage of users (or paying customers) who leave during a period. It's the opposite of retention. High churn means people are walking out the back door.
• Conversion rate — the percentage who take a step you want them to take. Visitors who sign up. Free users who go paid. Whatever the desired action is, what share did it.
• Engagement — how much and how often people use the product, measured in things like sessions or key actions taken.
• LTV vs. CAC — LTV (Lifetime Value) is the money a customer brings in over their whole time with you. CAC (Customer Acquisition Cost) is what you spent to win them. You want LTV comfortably bigger than CAC, otherwise you lose money on every customer you sign up.

You won't calculate these by hand. Tools do the arithmetic (that's Topic 13). The skill is knowing what each one means and reaching for the right one when a question lands on your desk.

Lesson 9.5 — One number the whole team points at

Dan from sales wants to chase enterprise logos. Priya wants to simplify the sign-up flow. The CEO has a new idea about an "AI assistant" every other week. Five smart people, five directions.

What keeps a team from pulling apart is a shared answer to one question: if the product is working, what number goes up?

That number is the North Star metric, the single measure that best captures the real value your product delivers. For a messaging app it might be "messages sent between friends." For a rental marketplace, "nights booked." For Lumi, something like "people who set a weekly budget and check it." None of those are dollars. A good North Star tracks real customer value rather than money, because money tends to follow value rather than lead it.

What does the work here is shared agreement, not the exact wording of the metric. Everyone holds one clear definition of success, so daily decisions point the same way instead of scattering.

For Maya, a North Star is a fast filter for any idea, including the CEO's: will this move it? It also guards against a sneaky failure where you juice one small metric and quietly hurt the product. Boosting sign-ups with spammy tactics looks great on that one chart while retention rots underneath. A well-chosen North Star keeps the whole team honest about what actually matters.

Worked example — Reading the numbers behind a "win"

Lumi runs a marketing push, and the team is thrilled: sign-ups doubled in a week. Dan is already talking about doing more of it. The CEO wants to triple the budget.

Maya doesn't celebrate the headline. She reads the funnel instead.

Acquisition, yes, is way up. But she looks one stage down and activation has fallen, fewer new users are reaching the "aha" moment, and week-2 retention has dropped too. The push pulled in the wrong people, who arrived, found nothing useful, and left. When she checks the North Star ("weekly active users doing the core action"), it hasn't moved at all. The doubled sign-ups were a vanity metric wearing a party hat.

Because she paired her metrics and read the funnel instead of the big number, she can say something specific in the room. The bottleneck sits at activation, not acquisition. More marketing would just pour more strangers into a funnel that's leaking at stage two. She redirects the team's effort toward fixing the first-run experience.

The impressive number hid the real story. Funnel-thinking pulled it back out.