Google Analytics 4 — Launch

What for: the industry standard for web and product analytics — an event model, funnels and a deep link to Google Ads. Free.

What it is and why

Google Analytics 4 is the global standard for web and product analytics, built on an event model (everything is an event: a view, a click, a purchase), cross-platform (site + app) and deeply wired into Google's ad stack. It's free. Reach for it when you need industry-standard analytics, work with an international audience, or run Google Ads. For a Russian audience, Yandex.Metrica is usually more practical (access to GA from Russia is unreliable).

How to do it (step by step)

Create a GA4 account and property at analytics.google.com.
Install the tag. Via Google Tag or Google Tag Manager (recommended) — GTM makes managing events easier without touching code.
Understand the event model. GA4 has no "pageviews" in the old sense — everything is events and parameters. The basics are collected automatically (Enhanced Measurement).
Set up key events (conversions): mark generate_lead, purchase and form submissions as Key Events — otherwise you won't see conversions at all.
Funnels (Explore). Build funnels and user paths in Explorations — powerful, but it takes learning.
Link it to Google Ads for attribution and ad optimization.
BigQuery export (free in GA4) — dump raw events for deep analysis with no sampling.
Consent (Consent Mode) and a cookie banner — for 152-FZ and GDPR.

Ready-made recipe

One funnel across app and site: GTM installs GA4 on the site and in the app → the events view_item → add_to_cart → begin_checkout → purchase are collected the same way everywhere → in Explorations you build the funnel and see that 60% drop off at begin_checkout → the Google Ads link shows which campaign brings actual purchases (not clicks) → the raw data streams into BigQuery for a detailed breakdown. One picture across every platform.

What it costs (honestly)

The logic: free for standard use (with limits and sampling at large volumes). The paid GA360 is the enterprise tier with higher limits (expensive, for large companies). The BigQuery export is free (you only pay for BigQuery storage and queries above the free tier). These change.

When it fits / when it doesn't

Fits: an international audience; you need the industry standard; you work with Google Ads; you're cross-platform (site + app); you want a BigQuery export.
Doesn't fit: your audience is in Russia (GA access is unreliable — use Metrica); you want simplicity (GA4 is complex); you care about privacy and going cookie-free (Plausible); you want product analytics with session replay and feature flags in one place (PostHog).

Risks and responsibility

Access from Russia is unreliable — it may require workarounds, which is a continuity risk for Russian projects. 152-FZ/GDPR: cookies and personal data collection mean consent (Consent Mode) plus a policy; the data goes to Google's servers (cross-border transfer — weigh the risks for Russian personal data). Sampling distorts the numbers at large volumes. Ads require ORD labeling. See the section disclaimer.

🤖 Prompt helper

You are a GA4 analyst. The project is a "{site/app}", the audience is in {geography}, the goal is {conversions/advertising}. Help me: (1) decide which events and Key Events to set up for my funnel; (2) install it through GTM without extra code; (3) build the funnel in Explorations and read it correctly; (4) configure Consent Mode and cookies for 152-FZ/GDPR. Account for unreliable access from Russia and tell me whether I also need Metrica. Step by step.