When Universal Analytics was switched off in July 2023, most Indian marketers migrated to Google Analytics 4 by necessity — not by choice. Two and a half years later, the pattern I see repeatedly when auditing analytics setups for Kerala businesses is the same: GA4 is installed, GTM is connected, and the dashboard is open. But only the Overview, Realtime, and Sessions graph get any attention. The other 90% of the platform sits unused.
GA4 was rebuilt from the ground up around events rather than sessions, and that shift unlocks report types that UA could never produce. The twelve reports below are the ones that actually change marketing decisions — from understanding which WhatsApp campaigns drove conversions during Onam to identifying exactly which step in your contact form funnel loses 60% of visitors. Each section includes where to find the report in GA4's navigation so you can open it right now.
1. Traffic Acquisition — Your Starting Point for Every Campaign Review
Location: Reports → Acquisition → Traffic acquisition
GA4's Traffic Acquisition report groups sessions by their first channel — Organic Search, Paid Search, Organic Social, Direct, Referral, and others. For Kerala businesses running simultaneous Meta Ads, Google Ads, and WhatsApp broadcast campaigns, the critical addition over UA is the ability to break down the "Unassigned" and "Direct" categories by adding a secondary dimension of Session source / medium.
WhatsApp links shared without UTM parameters land in Direct traffic and inflate that category considerably. When a retailer in Kochi shared a sale link via WhatsApp Broadcast to 4,000 contacts during Onam 2025 without UTM tags, GA4 showed a 340% spike in Direct sessions — which looks like branded search intent but is actually campaign traffic. The fix: tag every WhatsApp link as utm_source=whatsapp&utm_medium=broadcast&utm_campaign=onam_sale_2025. Same approach applies to Telegram channel links. Once tagged properly, the Traffic Acquisition report separates these cleanly under a custom channel grouping you can create in Admin → Data Settings → Channel Groups.
2. Landing Page Performance — Where Organic Traffic Actually Converts
Location: Reports → Engagement → Landing page
This report shows which pages users land on first, with columns for Sessions, Engaged Sessions, Engagement Rate, and — crucially — Conversions. In UA, you had to cross-reference Goals with Landing Pages in a secondary dimension. In GA4, mark your whatsapp_click and contact_form_submit events as Conversions, and this report shows you directly which landing pages generate actual leads.
Filter the report by Session default channel group → Organic Search to see which blog posts or service pages are bringing in organic leads, not just traffic. A service page may attract 200 monthly sessions but generate 8 WhatsApp clicks, while a blog post with 800 sessions generates zero. That comparison justifies where to invest content effort next quarter.
3. Funnel Exploration — Diagnosing Where You Lose Visitors
Location: Explore → Funnel exploration (create new)
Funnel Exploration is one of GA4's genuinely powerful additions. You define steps using events or page paths and see the percentage drop-off at each step. For a service business, a typical funnel looks like: session_start → page_view (service page) → scroll_50 → whatsapp_click or contact_form_submit.
For an Ernakulam interior design firm I worked with, the funnel showed 68% of users who viewed the portfolio page never scrolled past 30% — meaning most of the best project images were below a fold nobody reached. Moving three high-quality images above the fold increased the scroll_50 rate by 41% and directly improved WhatsApp click volume. You cannot see this in any overview report — only in a custom Funnel Exploration.
Enable the Closed funnel toggle to count only users who entered at Step 1. Open funnels inflate numbers by including users who jump into the middle of the sequence from a different page.
4. User Explorer — Understanding Individual High-Value Leads
Location: Explore → Free form, then drag "App instance ID" or use Reports → User → User explorer
For businesses where a single client is worth ₹2–5 lakh, you do not want to make decisions based on averages. User Explorer lets you click into any specific user's journey — every page they viewed, every event they triggered, every session they had — in chronological order.
When a client sees a lead from an unusual source (say, someone from Dubai who viewed pricing three times across five days before calling), User Explorer reveals the full path: arrived via Organic Search on a Malayalam keyword → visited two service pages → returned directly two days later → scrolled to the FAQ → clicked the WhatsApp button. That sequence tells you the FAQ section is doing serious conversion work for high-consideration buyers, which is not visible in any aggregate report.
5. Cohort Analysis — Measuring Onam Campaign Retention
Location: Explore → Cohort exploration
Cohort Analysis groups users by the week or month they first visited, then shows what percentage return in subsequent weeks. For Kerala e-commerce businesses with a strong Onam peak, this report answers a question that matters: of all the new users who came during Onam week, how many came back in November? December?
In 2025, a Thiruvananthapuram gifting business acquired 1,800 new users in the two weeks before Onam. Cohort Analysis showed that only 6% returned within four weeks — compared to 22% for users acquired via organic search during the same period. This single insight redirected their retention budget toward email re-engagement for organic acquirees rather than broad retargeting of all Onam visitors.
You can also compare cohorts from different campaigns: set Cohort inclusion to a specific event like first_visit filtered by campaign name, and compare the Onam Meta Ads cohort against the Onam Google Ads cohort across 8 weeks of retention.
6. Path Exploration — Mapping What Users Do After the Homepage
Location: Explore → Path exploration
Path Exploration visualises branching user flows — what pages users visit after a starting point, or what pages they came from before a conversion event. Set the starting node to session_start to see the most common first pages after arrival, or set it to a conversion event and work backward to see which pre-conversion paths are most common.
A practical use: start from page_view on your homepage and see which second page gets the most traffic. If 40% go to Blog and 25% go to Services, but the Services visitors convert at 8x the rate, your homepage navigation may be pushing too many users toward content rather than toward commercial intent pages.
For cart-based sites, set the endpoint to begin_checkout and look backward to see which pre-cart pages appear most. Often a specific product comparison page or FAQ appears in 60–70% of checkout paths — that page deserves prominent internal linking.
7. Campaign Attribution — Data-Driven vs Last Click
Location: Advertising → Attribution → Model comparison
By default, GA4 uses Data-Driven Attribution, which distributes conversion credit across all touchpoints based on Google's ML model. Last Click, by contrast, gives 100% credit to the final click before conversion. The difference matters enormously when evaluating whether your awareness campaigns (YouTube, Meta Reach campaigns) contribute anything to conversions.
Switch between models in the Model Comparison tool to see the same conversions re-attributed. Channels like Organic Social and Display typically gain credit under Data-Driven that they get zero credit for under Last Click. This is why Meta Ads often shows higher reported conversions than GA4 does — Meta counts view-through conversions and uses a 7-day click window, while GA4's Last Click gives it nothing if a Google Search click happened afterward.
For Kerala service businesses running Meta Ads for brand awareness alongside Google Ads for bottom-funnel search terms, the Attribution report usually shows that Meta deserves 15–25% of conversion credit under Data-Driven — enough to justify the spend even when GA4 Last Click reports show it generating "zero conversions."
8. Audience Segments — NRI Kerala Users and High-Value Purchasers
Location: Admin → Audiences (to create), then apply in Explorations
GA4 Audiences are reusable segments you can build once and apply across multiple reports. Two segments that are immediately useful for Kerala businesses:
NRI Kerala segment: Country = United Arab Emirates OR Saudi Arabia OR Qatar OR Kuwait, AND Language contains ml (Malayalam browser language) OR page_view URL contains /ml/. This isolates Gulf diaspora users who visit Kerala business sites, often researching services for family members back home. They typically have higher disposable income and longer consideration cycles.
High-value lead segment: Users who triggered whatsapp_click AND had session duration over 3 minutes AND viewed 3 or more pages. Apply this segment to the Traffic Acquisition report to see which channels bring in genuinely engaged visitors rather than bounced sessions.
Once created, Audiences can also be published to Google Ads for remarketing — the NRI Kerala segment is particularly effective for Diwali and Onam campaigns targeting Gulf-based buyers.
9. Event Tracking Audit — Fixing Silent Data Errors
Location: Reports → Engagement → Events
The Events report shows every event GA4 has recorded, with total counts and users. Use it monthly as a data quality check. Common errors that go unnoticed for months:
Duplicate form events: A Thank You page fires contact_form_submit, but GTM also fires it on a button click — resulting in 2x the true conversion count. Check if your conversion event count is suspiciously close to double the expected lead volume.
Missing events after a site redesign: If your developer changed a button's CSS class or ID, GTM's click-based trigger stops firing. The Events report will show whatsapp_click dropping from 45/month to 0 — a silent failure that makes campaigns look ineffective.
Event name mismatches: One developer tagged the event as contact_form_submit while GTM has a trigger looking for form_submit. The Events report shows both exist but the conversion is only tracking one. Standardise all event names in a naming convention document shared with your developer and GTM editor.
10. Geographic and Device Report — The Kerala Mobile Gap
Location: Reports → User → Tech and User → Demographics
Filter the Tech report to show Device Category with a secondary dimension of Engagement Rate and Conversions. For almost every Kerala service business I have audited, mobile generates 70–80% of sessions but 30–40% of conversions. Desktop users, who are fewer, convert at 3–4x the rate.
This gap usually has two causes: contact forms that are difficult to fill on mobile, and pages that load slowly on mid-range Android phones (the dominant device in Kerala's tier-2 and tier-3 districts). Use the Geographic dimension to see if Ernakulam users convert better than Wayanad users — which often reflects network speed differences as much as intent differences. Wayanad and Idukki users on slower 4G connections need pages under 3 seconds to load; GA4's engagement data shows whether you have a speed problem before your developer even runs a Lighthouse test.
11. Core Web Vitals Integration — GA4 + Search Console
Location: Reports → Acquisition → Search Console (after linking)
Link GA4 to Search Console under Admin → Product Links → Search Console Links. This adds a Search Console section under Acquisition with Google organic search traffic broken down by query, landing page, clicks, and impressions.
More useful for technical performance: create a custom Exploration using the engagement_time_msec metric alongside landing page. Pages with low engagement time (under 8,000 ms average — roughly 8 seconds) are often fast-loading pages users are bouncing from, or they indicate that GA4's enhanced measurement is firing page_view on JavaScript navigation before the page is usable. Cross-reference pages with high impressions in Search Console but low engagement time in GA4 — these are priority pages where content or speed improvements will directly affect both rankings and conversions.
12. Predictive Metrics — Pre-Onam Targeting with Purchase Probability
Location: Admin → Audiences → New Audience → Predictive
GA4's predictive audiences use Google's ML models to score users on Purchase Probability, Churn Probability, and Predicted Revenue. These are available only if you have enough conversion data (GA4 requires at least 1,000 returning purchasers and 1,000 non-returning purchasers over 28 days to activate the models).
For Kerala e-commerce businesses with seasonal spikes, the most actionable predictive audience is Likely 7-day purchasers — users who visited in the past 28 days and are predicted to convert within the next 7 days. Publish this audience to Google Ads and run a high-bid Performance Max or Search campaign targeting them in the two weeks before Onam, Vishu, or Christmas. You are spending on users already primed to buy, not trying to create intent from scratch.
Churn Probability is useful post-Onam: identify users who were active during the festival but have not returned in 30 days and run a re-engagement campaign before the next seasonal peak. A Kochi fashion retailer used this approach in November 2025 to recover 18% of lapsed Onam buyers ahead of their Christmas sale — at a cost-per-click significantly lower than cold prospecting campaigns.
FAQ
GA4 shows different numbers than Meta Ads and Google Ads — who is right?
Neither is wrong — they are measuring different things. GA4 uses session-based attribution starting from the first page load, while Meta Ads counts a conversion if the user clicked or viewed your ad within its default attribution window (7-day click, 1-day view). Google Ads uses 30-day click attribution by default. A user who clicked a Meta ad on Monday, returned via Google Search on Thursday, and converted — GA4 credits Organic Search, Meta Ads claims the conversion, and Google Ads may also claim it if a Search click happened during the window.
A 20–40% discrepancy between GA4 and Meta Ads is normal for Indian campaigns where users typically take 3–5 touchpoints before contacting a service business. To reconcile: compare identical date ranges, switch GA4 to a 30-day attribution window under Admin → Attribution Settings, and use the Campaign Attribution report to see how the Data-Driven model distributes credit across channels rather than treating any single platform's reported number as the truth.
How do I set up proper event tracking for a Kerala service business?
Five custom events cover the lead journey for most Kerala service businesses. Configure all of them via Google Tag Manager rather than hardcoding in your site HTML. First, contact_form_submit — triggered on the Thank You page URL or on a form confirmation event from your CMS. Second, whatsapp_click — triggered by a Click URL contains wa.me trigger, since WhatsApp is the primary lead channel for local clients across every Kerala district. Third, phone_click — triggered on tel: link clicks, which matter especially for older demographics in districts like Kottayam, Thrissur, and Palakkad. Fourth, scroll_50 — fired when users scroll 50% of a service page, signalling serious reading intent that precedes conversion. Fifth, video_play — if you have YouTube embeds or client testimonial videos on your site.
Once these events appear in GA4, mark whatsapp_click, phone_click, and contact_form_submit as Conversions under Admin → Events. This makes them available in all conversion columns across every standard report without creating separate Goals.
Should I enable the GA4 BigQuery export and what does it cost?
The BigQuery export itself is free — GA4 charges nothing to push your raw event data to BigQuery. You pay only for Google Cloud storage and query costs. For a Kerala SMB site processing around 1 million events per month, storage runs under ₹200/month. If you run 5–10 SQL queries per week for custom reporting, query costs typically stay under ₹300/month — well under ₹500/month in total.
The practical use cases that justify enabling it: building custom funnel queries that GA4's UI does not support natively, blending GA4 event data with CRM exports to see which lead source produces the highest client lifetime value, running raw retention queries on event-level data, and building Looker Studio dashboards from SQL rather than from the GA4 connector (SQL-based dashboards refresh significantly faster). To enable: GA4 Admin → BigQuery Linking → link your Google Cloud project → choose Daily or Streaming export. Streaming costs slightly more on the BigQuery side but gives you intraday data access, which is worth it if you are monitoring live campaign performance during high-stakes sale days like Onam or a product launch.
Closing Note
GA4 rewards the marketer willing to go past the Overview screen. The twelve reports above are not advanced features reserved for enterprise teams — they are available on every free GA4 property. The difference between businesses that get useful data from GA4 and those that don't is almost never the tool; it's the combination of clean event tracking, a habit of checking the right reports, and the willingness to act on what the data shows.
If your GA4 setup has accumulated months of mis-tagged events, duplicate conversions, or missing WhatsApp click tracking, fixing the data layer first will make every report on this list immediately more useful. Get that right before building dashboards or drawing conclusions from the numbers.