GA4 Migration

Your GA4 setup is probably broken. Let's verify.

UA was sunset in mid-2023. Most organizations made the switch. Far fewer organizations made the switch correctly. In 2026, the most common GA4 issues we encounter are: duplicate conversion events from parallel GTM containers, bounced sessions not counted because of deferred gtag.js loaders, enhanced measurement double-counting scroll and video events, and broken dataLayer variables that render as literal strings in GTM conditions. We audit every layer and fix what is wrong.

Audit My GA4 Setup

See All Analytics Services

Average broken configurations per GA4 audit

Conversion count gap we found at one client (GA4 vs. key events)

100%

Before/after verification on every fix

4wk

Typical audit to verified fix timeline

What We Audit and Fix

Six GA4 failure modes we diagnose and correct.

Event Parity Audit

We build a side-by-side comparison of every conversion event you tracked in Universal Analytics against what is currently firing in GA4. Gaps in event parity mean you are making decisions on incomplete data.

UA goal list vs. GA4 conversion event mapping
Event volume comparison by month to identify gaps
Missing event identification with GTM fix plan
Documentation of each event's firing condition and trigger

Double-Counting Diagnosis

The most common GA4 error we see: a site with two GTM containers, a gtag.js snippet, and a GA4 tag in each container. A single page load sends the GA4 measurement ID three times. Every event appears two or three times in reporting, making conversion rate and funnel data meaningless.

Network request audit to identify all GA4 payloads on page load
GTM container inventory: which containers are active on which pages
Duplicate tag identification and removal plan
Post-fix verification against expected event volumes

dataLayer Variable Validation

Many sites push variables to the dataLayer (page type, user status, product ID) so GTM tags can fire conditionally. If the variable names or value formats changed during a site rebuild, GTM conditions that depend on them break silently. Tags fire everywhere or nowhere instead of following the intended logic.

dataLayer variable inventory across all page types
GTM trigger condition audit against actual dataLayer values
Broken variable identification and fix specification
Testing protocol across all page templates

Enhanced Measurement Review

GA4's enhanced measurement fires automatic events for scrolls, file downloads, outbound clicks, and video engagement. When combined with manually implemented events tracking the same interactions, you get doubled counts. We audit which enhanced measurement events are active and which overlap with manual implementations.

Enhanced measurement settings audit per GA4 property
Overlap identification between automatic and manual events
Deduplication strategy and implementation
Impact assessment: how much did double-counting inflate your numbers

Deferred Loader Issue Fixes

Some site implementations load GA4 or GTM using a deferred script or behind a consent management interaction. Sessions where users bounce before any interaction never get counted. This artificially deflates session counts and inflates average session duration.

Script load order audit across all page templates
Bounce session undercount estimation
Consent management and GA4 initialization sequence review
Fix implementation for immediate (non-deferred) GA4 initialization

Conversion Event Mapping and Verification

GA4 distinguishes between events and key events (conversions). We map your business conversion events (form submissions, demo requests, trial sign-ups, purchases) to GA4 key events with proper counting method and value parameters, then verify in DebugView before declaring the migration complete.

Business conversion event to GA4 key event mapping
Value parameter configuration for revenue-bearing events
DebugView verification for every key event
Funnel visualization setup in GA4 Explore

How We Run the Audit

Four steps from suspected issues to verified clean setup.

Step 01

Full GA4 configuration audit

We review every component of your GA4 setup: the property configuration, all active data streams, every GTM container loading on the site, the full tag and trigger inventory, the dataLayer implementation, and the consent management setup. We document every issue we find, ranked by impact on data quality. Most audits surface three to six distinct problems.

Step 02

Fix specification and approval

For each issue found, we write a specific fix specification: exactly what to change in GTM, what to update in GA4, and what the expected outcome is. You approve the fix list before we touch anything. For clients with development teams, we coordinate the fixes with whoever owns the site codebase. For clients using tag managers only, we implement fixes directly.

Step 03

Implementation and DebugView verification

We implement fixes in GTM preview mode, verify every event in GA4 DebugView before publishing, and run a staging-to-production verification sequence. Nothing publishes until we have confirmed the fix works and has not introduced new issues. For conversion events, we verify against known test submissions to confirm end-to-end data flow.

Step 04

Post-fix monitoring and documentation

We monitor GA4 event volumes for two weeks after implementation to confirm that the fixed numbers are stable and within expected ranges relative to pre-fix baselines. You receive a documentation package covering every change made, every verification step completed, and the expected data ranges for your key events going forward.

GA4 Migration in Practice

The X0PA case: dual GTM containers, deferred loaders, and a 2x conversion inflation.

The Challenge

An AI-driven HR technology platform had a GA4 setup with two active GTM containers: one loading immediately on page load, one loading deferred behind a user interaction. They also had a standalone gtag.js snippet. The result was that every conversion event appeared 2 to 3 times in GA4 reports. Their November 2025 paid search showed 1,131 conversions in a month where the true number was closer to 200. The inflated data was driving paid spend decisions on false conversion rates.

Our Solution

We audited the full container and tag inventory, identified the three-way GA4 initialization (two GTM containers plus gtag.js), documented which tags were duplicated across containers, and built a fix plan: consolidate to one GTM container, remove the gtag.js snippet, and move the GA4 tag to immediate (non-deferred) initialization. We also identified that multiple dataLayer variables were pushing literal template strings instead of resolved values, breaking 4 GTM trigger conditions.

Results Achieved

1,131

GA4 conversion count before fix

November 2025 paid search (inflated)

~200

Actual conversion count after deduplication

Same period, real number

Broken GTM trigger conditions fixed

All caused by literal dataLayer variable strings

31%

Paid channel optimization improvement after real data available

Better CPL from accurate conversion data

FAQ

GA4 migration questions

Yes. UA was sunsetted in July 2023 and stopped processing data in July 2024. Despite this, a significant number of organizations that migrated to GA4 did so under time pressure with minimal QA. The result is that many GA4 setups are technically running but producing unreliable data due to double-counting, broken event mappings, deferred loading issues, or incorrect conversion event configuration. The migration may be complete in name; it is not necessarily complete in data quality.

Check two things. First, open your GA4 property and compare the "conversions" count to the "key events" count in the same report. If conversions is more than 20% higher than key events for the same time period, you likely have duplicate event firing. Second, compare your GA4 session count to your server log session count for the same period. If GA4 shows significantly fewer sessions than your server logs, you likely have a deferred loading issue missing bounce sessions. Both checks take 10 minutes and are worth doing before trusting any GA4 reporting.

Some developers optimize page performance by loading third-party scripts (including GA4 or GTM) with a defer attribute or behind a user interaction trigger (first scroll, first click). This improves initial page load speed but means that any user who bounces before the trigger fires never gets counted as a session. For content sites with high bounce rates, this can suppress session counts by 15 to 40%, making your organic bounce rate and average session duration metrics meaningless.

The audit itself takes one to two weeks depending on site complexity. A single GTM container with a straightforward tag structure takes 5 business days to audit. Multiple containers, custom dataLayer implementations, and consent management complexity can extend this to 10 to 15 days. Implementation and verification take an additional one to two weeks. Total timeline from start to verified clean GA4 setup is typically three to four weeks.

It depends on the issue. Most GA4 and GTM configuration fixes can be made through the tag manager interface alone, which requires Editor access to the GTM container. Fixes to the dataLayer implementation (adding or changing variables pushed in the site code) require developer access. We specify exactly which fixes need developer time versus GTM-only changes in our fix specification document before any implementation begins.

Ready to Verify Your GA4 Setup?

Book a GA4 audit before making any more data-driven decisions.

We will tell you within two weeks whether your conversion data is trustworthy. Most clients are surprised by what we find.

GA4 property and GTM container access required
Audit report delivered within 10 business days
Fix specification included at no extra cost