How to Set Up a Growth Analytics Stack from Scratch

You can't grow what you don't measure. That much is obvious. What's less obvious is that 80% of SaaS products are measuring the wrong things. They track page views and sessions while missing the behavioral events that actually predict growth — activation, engagement depth, feature adoption, and the specific friction points where users give up.

Setting up a proper growth analytics stack is not a 3-month project. Done correctly, the foundation is live in a week and generating actionable insights in two. Here is how to build it.

The 3-Layer Analytics Stack

A complete growth analytics stack has three distinct layers, each answering a different set of questions. Most teams have the first layer and are missing the second and third — which is exactly why their growth decisions feel like guesswork.

Website Analytics: GA4 or Plausible

Website analytics tracks what happens before the user enters your product. Which acquisition channels drive the most traffic? Which pages convert visitors into sign-up clicks? Where do paid and organic audiences come from?

Google Analytics 4 is free, deeply integrated with Google's ad ecosystem, and industry-standard. It is complex to configure correctly but powerful once set up. If you run any Google Ads or need multi-touch attribution, GA4 is the default choice.

Plausible is a privacy-first alternative that is simpler, lighter, and GDPR-compliant without cookie banners. Its dashboard is immediately readable. If you are not running paid acquisition at scale, Plausible gives you 90% of what you need with 20% of the setup complexity.

Whichever you choose, configure these events from day one:

Sign-up click (button click on the marketing site)
Sign-up form submission
Plan page view
Pricing CTA click
Demo request or contact form submission

Product Analytics: PostHog or Mixpanel

Product analytics tracks what happens inside your product. This is the layer most teams are missing, and it is the most valuable one for growth decisions.

PostHog is open-source, can be self-hosted, includes session recording alongside analytics, and is increasingly the default for early-stage SaaS. Its free tier covers most pre-Series A products comfortably. The autocapture feature means you start collecting data immediately, before you have event tracking implemented.

Mixpanel is more mature, has superior funnel and cohort analysis, and scales better for high-volume products. It is also more expensive and more complex to set up correctly. For products with 10,000+ monthly active users and a dedicated analyst, Mixpanel is the better long-term platform.

For most early-stage SaaS: start with PostHog, plan a migration path to Mixpanel when you need it.

Revenue Analytics: Stripe + ChartMogul

You need to see revenue in the context of cohorts, channels, and product behavior. The Stripe dashboard alone is insufficient — it shows you totals, not trends.

Stripe's native dashboard gives you MRR, churn, and transaction history. It is enough to start.

ChartMogul (or Baremetrics, or ProfitWell) connects to Stripe and gives you MRR growth, MRR churn rate, net revenue retention, LTV by cohort, and churn analysis by plan and billing period. These are the metrics that investors ask about and that reveal the health of your business below the surface.

Set up ChartMogul when you have 20+ paying customers and need to understand retention and expansion at the cohort level.

The 10 Events Every SaaS Should Track

These are the behavioral events that matter most for growth analytics. Implement all 10 from launch:

user_signed_up — Immediately on account creation. Properties: acquisition channel, referrer, plan type.
onboarding_step_completed — One event per onboarding step. Property: step name.
activation_event — Your product's defined aha moment action. This is the single most important event in your system.
feature_used — Core feature interactions. Property: feature name.
invite_sent — If you have any sharing or collaboration feature, track every invite.
upgrade_initiated — When a user clicks to upgrade. Property: current plan, target plan.
payment_completed — On successful subscription or one-time payment.
payment_failed — On failed charge attempts.
session_started — On every authenticated session start. This is how you calculate DAU and WAU.
user_churned — On account cancellation or subscription cancellation. Property: reason (from exit survey), plan, time since sign-up.

For each event, add consistent properties: user_id, account_id, plan, created_at (user's sign-up date). These enable cohort analysis across your full event history.

Building Your Core Dashboards

Acquisition Dashboard

This dashboard answers: where are your users coming from, and which sources produce the best users?

Metrics to include:

New sign-ups by source (week over week)
Activation rate by source
Day-7 retention by source
Top landing pages by conversion rate

The activation rate and retention columns are what transform this from a vanity dashboard into a decision-making tool. A source that drives 200 sign-ups at 10% activation rate is less valuable than one that drives 50 sign-ups at 50% activation rate.

Activation Funnel

This dashboard answers: where in the onboarding journey are users dropping off?

Build a step-by-step funnel from sign-up to activation event. Every step should show the number of users who entered that step, the percentage who completed it, and the average time to complete. The step with the biggest drop is your highest-priority UX fix.

Update this funnel every time you make a change to onboarding. The funnel is your onboarding scorecard.

Retention Cohorts

This dashboard answers: do users who sign up in a given week return and stay?

A retention cohort table shows, for each cohort of users who signed up in a given week, what percentage were still active at Week 1, Week 2, Week 4, Week 8, and Week 12.

A healthy SaaS product shows a curve that flattens — early drop-off followed by a stable retained core. A product with product-market fit problems shows a curve that keeps declining toward zero. If your Week-8 retention is below 15%, growth spending will not fix your underlying retention problem.

The Metrics That Actually Predict Growth

Many teams optimize for metrics that feel good but don't predict anything. The metrics that actually predict growth are:

Activation rate — If it's below 30%, growth spending before improving activation is a waste.

Day-7 retention — The strongest leading indicator of Month-1 retention. Benchmark: 30%+ for consumer, 40%+ for B2B.

Product-qualified leads (PQLs) — Users who have hit a specific engagement threshold (e.g., used the product 5+ times, or completed a meaningful workflow) and are on a free plan. PQLs convert to paid at 3–5x the rate of sign-ups who haven't hit that threshold.

Net revenue retention (NRR) — The percentage of revenue from existing customers retained after accounting for churn, downgrades, and upgrades. NRR above 100% means expansion revenue offsets churn — you grow even without adding new customers. This is the most powerful metric in SaaS.

Common Analytics Mistakes

Tracking too many events without a taxonomy. If your event log has 200 different event names after 3 months, you can't trust the data. Define a naming convention (object_action, e.g., project_created) and enforce it.

Not backfilling acquisition source. If you don't capture the acquisition channel at sign-up time, you can never connect product behavior to acquisition source later. Capture UTM parameters on sign-up and store them on the user record.

Ignoring mobile. If your product has a mobile app or a responsive web experience, your analytics must be mobile-aware. Separate mobile and desktop behavior — they are different experiences with different conversion patterns.

Optimizing activations while ignoring retention. Activation rates without retention data are meaningless. A product that activates 60% of users but retains none of them past Week 2 is not a growth problem — it is a product problem.

An analytics stack built correctly from the start saves months of guesswork and prevents the classic mistake of spending acquisition budget on a product that doesn't retain the users it earns.

I'm Mehdi Yatrib, a growth consultant based in Casablanca. I help SaaS teams implement their analytics foundation, instrument the right events, and build the dashboards that support real growth decisions.

Book a growth consulting call at yatrib.me