GA4 for SEO: The Four Views That Matter
Search Console ends at the click — it can tell you someone chose your result, and nothing after. Google Analytics 4 picks up exactly there: what visitors did on your site, whether they engaged, and whether they became enquiries and customers. GA4 is also a maze of reports most of which an SEO never needs. This lesson gives you the four views that matter and permission to ignore the rest.
The Division of Labour
| Search Console (Lesson 8.1) | GA4 (this lesson) | |
|---|---|---|
| Territory | Before and at the click | After the click |
| Answers | Which queries, which positions, which pages appear, who clicks | What visitors do, how engaged they are, what converts |
| Source | Google’s search logs | A tracking tag on your site (all traffic sources, not just Google) |
| SEO’s use | Diagnose rankings and visibility | Prove value and find conversion leaks |
Setup notes, kept short: install the GA4 tag (most WordPress setups do this via a plugin or a snippet in the header tool), and immediately do two things — link GSC to GA4 (Admin → Product links → Search Console links; unlocks query data inside GA4’s reports) and define your key events (below). Data starts at installation, so like GSC: set up today, analyse later.
First: Isolate Organic (The Skill Behind Every View)
GA4 lumps all traffic together by default; every SEO question starts by isolating search traffic. The core concept is session default channel group — GA4’s classification of where sessions came from. The one that matters here: Organic Search. In any report, add the filter or dimension for channel group = Organic Search, and you’re looking at SEO’s slice alone. (Adjacent channels worth recognising: Direct includes people typing your URL — partly the brand demand from Lesson 5.3; Organic Social, Referral and Paid each have their own stories, none of them yours to claim in an SEO report.)
Engagement Rate: The Metric That Replaced Bounce Rate
Old-Analytics veterans ask about bounce rate; GA4 rethought it. An engaged session is one that lasted 10+ seconds, or triggered a key event, or viewed 2+ pages. Engagement rate = engaged sessions ÷ total sessions — and it’s a far better content-quality signal than bounce ever was, because a visitor who read your entire answer for 3 minutes and left satisfied was a “bounce” under the old definition and an engaged session now.
How to read it, calibrated to what you learned in Lesson 3.3: engagement rate is a proxy for “did the search end on this page?” A landing page with organic engagement far below your site’s norm is either attracting the wrong intent (Lesson 2.1 mismatch), loading badly (Lesson 4.4), or failing the answer-first structure — three diagnoses, all covered, and this metric tells you where to look.
Key Events: Making GA4 Speak Business
GA4’s “key events” (its term for conversions) are the bridge from traffic to value — and for Indian small-business sites, the events that matter are rarely purchases:
- WhatsApp link clicks — for most Indian service businesses, the conversion. A click on your wa.me link is an enquiry.
- Phone number taps (tel: links) — the mobile-first equivalent.
- Form submissions — contact and order forms.
- Purposeful downloads — the lead-magnet PDFs and templates (your Lesson 5.1 tools doing measurable work).
- E-commerce purchases where applicable.
Mechanically: GA4’s enhanced measurement auto-captures outbound clicks, downloads and more as events; in Admin → Events you mark the ones that represent enquiries as key events (click events on wa.me/tel: links may need a small setup via your tag or a plugin — a one-time job worth doing properly, because everything in Lesson 8.3 reports against these).
The Four Views an SEO Actually Needs
GA4 has dozens of reports and an Explore builder that can do anything. You need these four, monthly:
The Traps: Where GA4 Wastes SEO Time
- Report tourism: demographics, tech reports, real-time dashboards — interesting, rarely actionable for SEO. The four views answer the questions that change decisions; schedule them monthly and resist wandering.
- Comparing GA4 numbers to GSC numbers: clicks (GSC) and sessions (GA4) are defined and counted differently — they will never match, and the gap means nothing. Use each tool for its territory.
- Small-number theatre: engagement rate on 20 sessions, conversion swings of 3 vs 5 — noise wearing a percentage. Set minimum sample habits (≈100 sessions before reading rates) and let months, not days, tell trends — the same patience discipline as Lessons 4.4 and 5.4.
- Consent and blocking reality: GA4 sees somewhat less than everything (ad blockers, consent declines). Fine for trends and comparisons — the SEO use — just never treat its absolute counts as a census.
Key Takeaways
- Clean division: GSC owns before-the-click, GA4 owns after — link them once, and never expect their numbers to match.
- Every SEO question in GA4 starts with isolating Organic Search via channel group.
- Engagement rate replaced bounce rate — and it operationalises Lesson 3.3: did the search end on this page?
- Key events make GA4 speak business: WhatsApp clicks, calls, forms, downloads — “47 enquiries from organic” beats “traffic up 30%” in every report ever written.
- Four monthly views suffice: organic trend, landing pages by key events, engagement outliers, path awareness — everything else is report tourism.
- Respect the noise floor: ~100-session minimums, month-scale trends, and GA4 as a trend instrument, not a census.