GA4 Common Mistakes: The 20 Errors That Break Properties (2026)

The 20 most common GA4 mistakes

Mistake 1 — Data retention left at 2-month default *(Most costly)*

Frequency: Extremely common — most new GA4 properties are never changed from default

Damage: All event-level data in Explorations older than 2 months is permanently inaccessible. Year-on-year comparison is impossible.

Symptom: Explorations only load data for the last ~60 days; date ranges beyond 2 months return no data

Fix: Admin → Data Settings → Data Retention → 14 months. Not retroactive — immediate action required on every new property.

Mistake 2 — page_view marked as a key event

Frequency: Very common

Damage: Key event count is meaninglessly inflated. "Conversion rate" appears to be 100%+ because every session has at least one page_view.

Symptom: Key event count equals or exceeds session count; key event rate shown as ~100%

Fix: Admin → Events → find page_view → toggle Key event OFF

Mistake 3 — Consent Mode V2 not implemented on EU/UK sites

Frequency: Common (especially smaller agencies and in-house teams)

Damage: Google Ads modelled conversions don't activate; conversion undercounting 20–50% in EU/UK markets; technically non-compliant with March 2024 Google Ads requirement

Symptom: gcd parameter absent from GA4 network requests; modelled conversions absent from Google Ads reports

Fix: Implement Consent Mode V2 via CMP GTM template (Cookiebot, OneTrust, Usercentrics)

Mistake 4 — Duplicate transaction IDs

Frequency: Common

Damage: Revenue over-reported (sometimes 2x or more); Smart Bidding corrupted by inflated conversion data

Symptom: GA4 revenue significantly exceeds platform revenue; same transaction_id appears multiple times in BigQuery

Fix: Implement sessionStorage deduplication; use checkout_completed (Shopify) or server-side _ga4_tracked meta flag (WooCommerce)

Mistake 5 — Purchase event fires on page load instead of order confirmation

Frequency: Common in GTM implementations where trigger is "Page URL contains /thank-you"

Damage: Every user who lands on the thank-you URL (including returns via bookmark, email link) fires a purchase event. Revenue massively overstated.

Symptom: GA4 transactions significantly exceed platform orders; each transaction_id appears 5–10 times

Fix: Fire purchase event from a custom event triggered by the server confirming the order (checkout_completed API, server-side dataLayer push on first load only)

Mistake 6 — No internal traffic filter

Frequency: Very common

Damage: Every session from the office, agency, and developer machines inflates session counts and conversion rates. For small teams, internal traffic can represent 10–40% of reported sessions.

Symptom: Unusually high conversion rates; site appears busier than expected; developers' test purchases inflate revenue

Fix: Admin → Data Streams → Define internal traffic → add IPs → create Data Filter to exclude

Mistake 7 — UTM parameters on Google Ads URLs (conflicting with auto-tagging)

Frequency: Common

Damage: Manual UTM parameters override the gclid attribution. Google Ads sessions are misattributed in GA4; Google Ads conversion tracking breaks.

Symptom: Google Ads shows impressions and clicks; GA4 shows no google/cpc sessions; conversions in Google Ads don't match

Fix: Use auto-tagging only for Google Ads. If manual UTMs are needed alongside gclid, ensure the tracking template format preserves gclid: {lpurl}?utm_source=google&gclid={gclid}

Mistake 8 — items array empty or missing on purchase event

Frequency: Common in custom implementations

Damage: Product-level e-commerce reporting unavailable; Google Shopping campaign optimisation broken; item_id ≠ Merchant Center product ID alignment fails

Symptom: GA4 → E-commerce Purchases shows (not set) for item names; Google Ads Products tab shows zero conversions at product level

Fix: Verify dataLayer push includes populated ecommerce.items array with item_id, item_name, price, quantity on the order confirmation page

Mistake 9 — Custom dimensions not registered in GA4 Admin

Frequency: Very common

Damage: Custom parameters are collected but invisible in standard reports and Looker Studio. Analysts can't use them without BigQuery.

Symptom: Attempting to add a custom parameter as a dimension in GA4 reports shows it's not available; the data exists in BigQuery but not in the UI

Fix: Admin → Custom Definitions → Custom Dimensions → + Create → register immediately when implementing new parameters

Mistake 10 — Enhanced measurement form tracking conflicts with custom form events

Frequency: Common

Damage: Every form submission fires 2 events — one from enhanced measurement, one from custom GTM tag. form_submit count is 2x actual.

Symptom: form_submit count in GA4 is approximately double the number of actual form submissions

Fix: Disable enhanced measurement form interactions (Admin → Data Streams → Web → Enhanced measurement → uncheck Form interactions) if custom form tracking is implemented

Mistake 11 — Agency GA4 property created in agency's own account

Frequency: Common in smaller agencies

Damage: When agency relationship ends, client loses access to all historical data. Data ownership stays with the agency.

Symptom: Client cannot log in to see their own GA4 data; agency is the sole account administrator

Fix (preventive): Always create client properties in the client's own Google Analytics account, with agency invited as Editor

Mistake 12 — Referral exclusions missing for payment gateways

Frequency: Common

Damage: Purchases appear attributed to PayPal, Stripe, or other payment processors as referral sources. Channel attribution for conversions is wrong.

Symptom: GA4 shows significant sessions and conversions attributed to paypal.com, stripe.com, or similar referral sources

Fix: Admin → Data Streams → Configure your domains → add payment gateway domains (paypal.com, stripe.com, secure.klarna.com, etc.)

Mistake 13 — wrong timezone set for the property

Frequency: Common

Damage: Day boundaries are wrong — sessions split between two days at the wrong clock time. Daily and weekly reports show incorrect patterns.

Symptom: Day-over-day traffic patterns look unusual; Monday spikes that should be Sunday evening traffic

Fix: Admin → Property Settings → Reporting time zone → set to business operations timezone

Mistake 14 — Sending PII in event parameters or page URLs

Frequency: Less common but high severity

Damage: GDPR violation (data minimisation), Google Analytics ToS violation, potential data breach notification obligation

Symptom: Email addresses or phone numbers visible in page_location values in GA4 reports or BigQuery

Fix: Enable GA4's URL redaction; implement GTM variable sanitisation to strip PII query parameters before they reach the GA4 tag

Mistake 15 — Audience membership duration too long for urgency campaigns

Frequency: Common in paid media teams

Damage: Cart abandonment audiences include users who abandoned 180 days ago — these users have near-zero intent. Ad spend wasted on cold audiences misidentified as hot.

Symptom: Cart abandonment remarketing shows high click volume but very low conversion rates

Fix: Set cart abandonment audience membership to 7–14 days maximum. Use longer durations only for win-back or look-alike seed audiences.

Mistake 16 — Exploration date range extended beyond event retention window

Frequency: Common (especially after switching from 2-month to 14-month retention)

Damage: Explorations show no data for the period before retention was extended, even though some data may exist.

Symptom: Funnel exploration shows zero sessions for periods older than 2 months (before retention was extended)

Fix: Understand that the change is not retroactive. For historical analysis, use BigQuery export if it was enabled before the retention change.

Mistake 17 — Marking micro-conversions as primary conversions in Google Ads

Frequency: Common

Damage: Smart Bidding optimises toward add_to_cart or begin_checkout (which are plentiful and low-commitment) instead of purchase (which is the actual business outcome). CPA appears low but revenue doesn't improve.

Symptom: Smart Bidding reports excellent "conversion" numbers; actual revenue or final purchase rate doesn't improve in proportion

Fix: Set purchase (with value) as Primary conversion in Google Ads. Set micro-conversions (add_to_cart, begin_checkout) as Secondary only — they inform bidding but don't set the goal.

Mistake 18 — User-ID containing raw email addresses

Frequency: Less common, high severity

Damage: GDPR violation (sending PII to Google); Google Analytics ToS violation; potential data breach

Symptom: User Explorer shows email addresses as user IDs

Fix: Hash User-ID with SHA-256 before sending to GA4. Never use raw email, name, or phone as User-ID.

Mistake 19 — Direct traffic over 35% without investigation

Frequency: Common (symptom, not always noticed)

Damage: Attribution loss — campaigns that should be credited with traffic are not. Budget decisions based on understated channel performance.

Symptom: Direct channel consistently > 35% of sessions; paid campaign sessions seem lower than expected based on spend

Investigation: Check for broken UTMs, missing cross-domain tracking, consent mode issues

Mistake 20 — Never annotating GA4 after configuration changes

Frequency: Extremely common

Damage: Future analysts (or your future self) cannot explain data anomalies. Tracking changes look like traffic changes. Stakeholder trust eroded when unexplained metric shifts appear.

Symptom: Unexplained 20% session drop in October with no annotation to explain that the internal traffic filter was added

Fix: Annotate within 48 hours of any tracking change, site change, or significant marketing event. 30 seconds per annotation pays dividends over years.