Product metrics answer a different question from "is the frontend fast?" — they ask whether the product delivers value: do people show up, do they reach the aha moment, do they come back, and do they stick around. Almost all of it is measured in the browser by frontend event tracking, but the numbers describe the product, not the frontend's health. Here's the set that actually changes decisions.
A note on scope. These are the metrics Measuring the Frontend deliberately hands off: product/growth metrics measured in the browser but owned by product. A frontend team's stake in them is instrumentation quality — events firing correctly, no double-counting, no gaps — not the numbers themselves. A slow or janky flow shows up here as drop-off, so watch them; just don't mistake them for a scorecard of the frontend.
Every metric below carries the tool that actually produces it, tagged by access tier: free
free tier or open-source (Google Analytics 4 (GA4), PostHog, Microsoft Clarity) · paid paid
product-analytics suites (Amplitude, Mixpanel, Heap, Pendo) · replay session-replay and
heatmap tools for the qualitative where and why (Hotjar, FullStory, Clarity) · pipe the
event pipeline / customer data platform (CDP) that feeds them all (Segment, RudderStack).
First principle: value delivered, not activity
The trap in product analytics is the vanity metric — a number that goes up and to the right and tells you nothing you can act on. Total registered users only ever climbs; it never tells you to do something differently tomorrow. The antidote is to ask, for every metric, "if this moves, does a decision change?" If not, it's a slide decoration.
The metrics below are grouped by the journey they describe — acquisition happens elsewhere, so product picks up the moment a user arrives: do they convert, do they engage, do they activate, and do they come back.
Conversion & funnel
Conversion measures whether visitors do the thing you want. A funnel breaks a multi-step flow into stages and shows where people fall out.
| Metric | Definition | Watch for | Tools |
|---|---|---|---|
| Conversion rate | % of visitors completing the goal action (signup, purchase) | Segment it — a blended rate hides that mobile converts at half of desktop | GA4, PostHog (free); Amplitude, Mixpanel (paid) |
| Funnel drop-off | Where in a multi-step flow users abandon | The single worst step is usually where the return on investment is | Funnel reports in Amplitude, Mixpanel, PostHog (paid/free); GA4 funnel exploration (free) |
| Micro-conversion | A smaller committed step (add-to-cart, email entered) | Leading indicator of the macro conversion below it | Event tracking in Amplitude, Mixpanel, PostHog (paid/free) |
| Macro-conversion | The primary business goal (purchase, paid signup) | The number the business actually cares about | GA4 conversions, Amplitude (free/paid) |
A useful distinction is micro vs macro conversions. The macro conversion is the goal (a purchase); micro-conversions are the committed steps on the way (viewed pricing, added to cart, entered email). Micro-conversions are leading indicators — they move before the macro number does, so they're where you spot trouble early.
Read a funnel by its biggest single drop, not its endpoint — if 40% start signup but only 28% finish, that 12-point leak is one form, and fixing it lifts everything downstream. The frontend's fingerprints are all over these steps (form user experience, latency, layout stability), which is exactly why a frontend team should watch the funnel even though it doesn't own the goal behind it.
Engagement
Engagement metrics measure how people interact once they're in. They sit close to performance, but they're direction-ambiguous — you have to interpret them against what the product is for.
- Bounce / exit rate — % of users leaving after a single page, and where they leave.
Usually reads as friction and dead ends, but a docs page that answered the question in one
visit is a "bounce" that's actually a win. Measure with: GA4 (
free), plus Hotjar or Microsoft Clarity (replay) to see where the leave happens. - Session duration & depth — time per session, and pages/actions per session. A longer
session is good for a game or social app and bad for a checkout flow or support tool,
where the goal is to get users done fast. Never assume "more time on site" is a win —
always ask what the product is trying to help the user do. Measure with: GA4 and any
product-analytics suite (
free/paid); FullStory (replay) for the session-level detail.
Activation, adoption, stickiness, retention
This is the core of product analytics — the arc from a new user's first value to a durable
habit. It's almost always captured by frontend event tracking — events usually flow through a
customer data platform (Segment, RudderStack pipe) into the analytics tools below — and the
instrumentation being correct is the frontend's real stake here.
| Metric | What it measures | Good direction | Tools |
|---|---|---|---|
| Activation rate | % of new users reaching the "aha" moment (the first real value) | Higher — it predicts everything downstream | Amplitude, Mixpanel (paid); PostHog (free) |
| Feature adoption | % of users who use a given feature | Depends — high for core, low is fine for niche | Pendo (paid); Amplitude, PostHog (paid/free) |
| Stickiness (DAU/MAU) | Daily actives ÷ monthly actives — how many days a month people show up | Higher for habit products; irrelevant for occasional-use ones | Native stickiness charts in Amplitude, Mixpanel (paid); PostHog (free) |
| Retention rate | % of a cohort still active N days/weeks later | Higher & flattening — a curve that hits a plateau means product-market fit | Cohort retention in Amplitude, Mixpanel, PostHog (paid/free) |
| Churn rate | % of users (or revenue) lost in a period | Lower — the mirror image of retention | Analytics above; Stripe billing (paid) for revenue churn |
Activation is the highest-leverage number in the list because it gates the rest — a user who never reaches first value can't be retained. Defining it well is the hard part: it's a specific early action correlated with long-term retention ("added 3 items", "invited a teammate", "sent first message"), not just "signed up".
Retention is best read as a curve, not a single number. Plot the % of each signup cohort still active over time. Three shapes tell the whole story:
| Curve shape | Verdict | What it means |
|---|---|---|
| Drops to zero | Users try it once and leave — no durable value | |
| Declines then flattens | The "smile" — a core keeps using it; the plateau height is your real retained base | |
| Rises back up | Rare — expanding value pulls dormant users back ("negative churn") |
The plateau matters more than the early slope: a curve that flattens at 40% has found a loyal base, while one that slides toward zero hasn't — however good the first-week numbers looked. Cohorting also isolates change: compare the users who joined after a redesign against those before, and you see whether the change actually moved retention.
Stickiness — daily active users over monthly active users (DAU/MAU) — compresses habit into one ratio: a DAU/MAU of 0.5 means the average user shows up ~15 days a month. It's a great headline for a daily-habit product (a chat app, a social feed) and a misleading one for a product people are supposed to use rarely (tax software, a real-estate portal) — don't hold every product to the same stickiness bar.
The North Star
Above the individual metrics sits one number a good product team keeps at the top: the North Star metric — the single measure that best proxies the durable value customers get from the product. It's chosen so that moving it means the product genuinely got more useful, not just busier.
| Product | North Star (proxy for value) |
|---|---|
| Marketplace lodging | Nights booked |
| Messaging | Messages sent between connected users |
| Collaboration tool | Weekly active teams |
| Media / streaming | Hours of content watched |
A North Star works only when it's paired with guardrails — counter-metrics that go bad if you chase the star too hard. "Messages sent" paired with nothing invites spam; paired with spam-report rate and retention, it stays honest. The pairing is what keeps a North Star from becoming just another number to game.
The North Star is a dashboard number — tracked in the same product-analytics tool (Amplitude,
Mixpanel paid) or modelled from raw event data in a business-intelligence (BI) layer (Looker,
Metabase, Mode).
This post covers the product slice of metrics measured in the browser. For the numbers the frontend directly owns, see Measuring the Frontend: Metrics That Actually Matter; for the demand side, see Measuring Marketing & SEO: Metrics That Actually Matter; and for the whole tour across delivery, reliability, and quality, see Measuring Software Engineering: Metrics That Actually Matter.