Why GA4 Fails at Conversion Tracking (and What to Do Instead)

The marketing world is obsessed with data — but are we measuring what matters? Google Analytics 4 promises enhanced tracking and deeper insight, yet several of its structural flaws can quietly sabotage your marketing with misleading conversion data, distorted audience metrics and a customer journey it can't fully see. This isn't about bashing Google. It's about knowing where GA4 misleads you, so you don't allocate budget on flawed numbers. Let's walk through the biggest flaws — and what to do instead.

THE DIRECT-TRAFFIC DELUSION

GA4 Attributes a large share of conversions to "direct traffic" — as if people simply materialized on your site and bought, like stumbling on an oasis in the desert. But people rarely convert on a whim, especially if you're not a household name. Most need several touchpoints, several nudges, multiple visits before they buy.

Filing those conversions under "direct" erases the weeks or months of marketing that actually moved the customer. It's like crediting the grocery-store cashier for the entire supply chain — ignoring the farmers, processors and distributors who did the real work. Was the conversion really "direct," or was it that abandoned-cart email, a retargeting ad or the blog post you labored over?

And the problem is getting worse, not better. Direct traffic is inflated by any break in the tracking chain — stripped referrers, missing UTMs, redirects that drop parameters, a tag missing on a landing page. In 2026 there's a new contributor too - referrer-less AI traffic. When someone arrives from ChatGPT or another AI tool that sends no referrer, GA4 has nowhere to file it — so it lands in "direct." A common rule of thumb is if direct traffic exceeds roughly 25% of your total, treat it as an attribution problem to investigate, not genuine type-in visits.

THE NEW-VS-RETURNING USER MYTH

GA4 Frequently miscategorizes returning visitors as brand-new ones. The mechanism is worth understanding precisely, because it's often blamed on GA4 alone when the real culprit is browser privacy. Safari's Intelligent Tracking Prevention caps JavaScript-writable first-party cookies at 7 days (and just 24 hours if the visitor arrived via a link decorated with cross-site tracking parameters). Once that cookie expires, GA4 loses the thread and counts the returning visitor as new. Brave applies similar limits.

Do you visit every site you frequent at least once a week? Of course not — life happens. But under these constraints, anyone who takes a break longer than the cookie window reappears as a "new" user. The result is inflated new-user counts, a distorted view of loyalty and audience size and campaigns built to chase "new" users who are actually returning ones. You can end up targeting the same people over and over, convinced they're fresh prospects. A quick diagnostic is to compare returning-user share by browser — if Safari shows far fewer returning users than Chrome for the same audience, ITP is collapsing your user IDs.

THE MISSING FIRST TOUCH

GA4 Stumbles hardest on the full customer journey because it so often loses the initial touchpoint and the first interaction is what set the whole journey in motion. It's like assembling a complex puzzle with several key pieces missing. Without knowing what first drew a customer to your brand, you can't accurately attribute the conversion or optimize what created the demand in the first place. Compounding this, Apple's App Tracking Transparency and ITP have punched large holes in browser-based tracking that no amount of GA4 configuration can patch. An iPhone user clicks your ad, browses, leaves and returns later via search to buy — and GA4 frequently fails to connect those sessions as one person.

WHAT THIS LOOKS LIKE IN REAL DATA

To make the gap concrete, here's data pulled directly from a real GA4 account and compared against a system that stitches data across sources, over the same retention period. It's a single client example, not a universal benchmark — but it's representative of the scale of distortion :

- Sessions (Aug 2023–Jul 2024) : GA4 reported 6.45 million; the stitched, deduplicated view showed 1.68 million.
- New users : GA4 reported 88.94%; the accurate view showed 37%.

When GA4 tells you nearly 89% of your users are new and the real figure is closer to 37%, every acquisition-vs-retention decision built on that number is skewed.

WHERE TO GO FROM HERE

Don't panic — these flaws aren't insurmountable. The fix is to stop relying on GA4's siloed, browser-limited view and stitch together a broader set of data:

- Your CRM (HubSpot or similar)
- Your email marketing platform
- Your ad platforms (yes — even Meta's own attribution adds useful context)
- GA4 itself, treated as one input among several rather than the source of truth

Combined, these let you build a far more accurate, people-based view of the journey - whether a source genuinely brings new customers, how touchpoints connect across channels and what your true ROI is.

This is exactly the gap Wicked Reports fills. Rather than trusting GA4 — or any single platform — to grade its own homework, it reconciles ad-platform, CRM and first-party sales data into one deduplicated, people-based view tied to real orders. That's how you see the actual first and assisted touchpoints driving revenue, instead of a pile of conversions dumped into "direct." See how it works in the platform overview, or book a demo to compare it against your own GA4 numbers.

GA4 might be the industry default, but "default" isn't the same as "accurate." Recognizing its flaws and complementing it with a true cross-channel source puts you a step ahead of every competitor still making budget decisions on distorted data.

FAQ

WHY IS "DIRECT TRAFFIC" MISLEADING FOR CONVERSIONS IN GA4?

GA4 files conversions under "direct" whenever it loses the original source — a stripped referrer, missing UTMs, an expired tracking cookie, or increasingly, referrer-less AI traffic from tools like ChatGPT. That erases the multi-touch marketing (emails, ads, content) that actually drove the sale, inflating "direct" and leading to inaccurate ROI and poor budget allocation. If direct exceeds ~25% of traffic, treat it as an attribution problem.

WHY DOES GA4 MISCOUNT NEW VS. RETURNING USERS?

The usual cause is browser privacy, not GA4 alone: Safari's Intelligent Tracking Prevention caps first-party tracking cookies at 7 days (24 hours with cross-site parameters), so returning visitors who've been away longer get recounted as new. This inflates new-user metrics and distorts audience size, loyalty, and retention measurement. Comparing returning-user share across browsers reveals the effect.

WHAT'S THE RECOMMENDED ALTERNATIVE TO RELYING SOLELY ON GA4?

Adopt an integrated, multi-source attribution approach that stitches together CRM data (e.g. HubSpot), your email platform, and your ad platforms (Meta, Google Ads) alongside GA4. This people-based view captures the full journey — the actual first and assisted touchpoints — for a far more accurate read on true ROI than GA4's browser-limited data can provide on its own.