Google Ads vs GA4: Why the Numbers Differ

If you have ever exported the same week from both tools and found two numbers that refuse to agree, you are not doing anything wrong. The mismatch is structural. This article walks through exactly what causes it, with a worked example you can map to your own store, and shows you which number to trust for which decision.

Why the numbers can never match

Google Ads and GA4 are not two views of one dataset. They are two separate measurement systems with different jobs.

Google Ads measures ad performance. It only ever counts interactions with Google ads, and it optimizes bidding toward the conversions it can claim. GA4 measures site behavior. It watches every session from every source and tries to describe the whole customer journey.

Because the questions differ, the answers differ. A gap somewhere in the range of ten to thirty percent between the two is common enough that it is not, on its own, evidence of a bug, according to Dataslayer's breakdown of Google Ads and GA4 discrepancies. What deserves an afternoon of investigation is a gap much larger than that. This same tension shows up across every tool in your stack, which is why it helps to have a single reference for reconciling your ecommerce data.

The five reasons your counts disagree

1. Different attribution models

Google Ads credits conversions using its own last-click logic within Google's channels, so a user who clicks your ad and later buys is credited to that ad. GA4 defaults to data-driven attribution, which splits one conversion's credit fractionally across every touchpoint in the journey.

So one real sale can appear as a full conversion in Google Ads and only a fraction of a conversion under GA4's "Paid Search" channel, with the rest handed to organic, email, or direct. If you want the deeper mechanics of how these models divide credit, our guide to attribution modeling tools breaks them down.

2. Click date vs conversion date

This one silently wrecks short date ranges. Google Ads records a conversion on the date of the ad click that earned it, not the date the purchase happened.

Say a shopper clicks your ad on a Monday and finally checks out the following Thursday. Google Ads books that conversion on Monday. GA4 books it on Thursday. Compare a single day and the totals can look wildly off even when the trailing two-week totals agree. Always compare on a window of a week or two, never a single day.

3. Conversion counting: "every" vs "one per click"

Google Ads lets you choose whether to count every conversion after a click or just one. For a purchase action set to "every," one shopper who buys three times after a single click counts as three conversions.

GA4 counts key events differently, typically deduplicating toward one per session depending on your setup. Same buyer, same behavior, two different totals — before any tracking loss even enters the picture.

4. GA4 loses events that Google Ads recovers

GA4 runs on browser-side JavaScript and cookies. Ad blockers, privacy browsers, closed tabs, and declined cookie consent all stop its tags from firing. Field estimates put the share of users affected by ad blockers and tracking prevention at roughly ten to twenty-five percent, per Audiense and Elevar's analysis of cross-tool data gaps.

Stacked together, those losses mean GA4's browser-based tags can miss a meaningful slice of real conversions, as Verde Media notes in its write-up of the GA4 and Google Ads discrepancy. Google Ads, meanwhile, leans on its own tag and Google's logged-in signals, so it recovers conversions GA4 never sees. That is a major reason Google Ads usually reports the higher number — the pattern we unpack in why Google Ads overreports compared to GA4.

5. Conversion modeling

When Google cannot directly observe a conversion — an opted-out user, a blocked tag — it estimates one with machine learning and reports the estimate as a counted conversion. GA4 exports the observed conversions, and Google Ads applies additional modeling on top, which can push the Ads figure above GA4's.

Modeled conversions are not fabricated demand. They are statistical estimates of real but unobservable sales. They can overshoot or undershoot, but they are an estimate of something that actually happened.

A worked example

Say your store runs one Google Ads campaign for a week. In reality, fifty people click an ad and eventually buy, at a forty-dollar order value. Here is how each tool reports that identical week.

Google Ads reports about fifty conversions. It credits every buyer who clicked, using last-click within Google's channels, and books each on the click date. A handful of Monday clicks that converted the next week even land in the prior reporting period. With "every conversion per click" on, a couple of repeat buyers nudge the count slightly higher.

GA4 reports about thirty-four conversions for Paid Search. Under data-driven attribution, roughly a dozen of those buyers first arrived via the ad but returned through organic or email before checking out, so GA4 splits their credit away from paid search. Then browser-side losses erase several more buyers entirely. Fifty real sales become thirty-four under the "Google / CPC" row.

Now put money on it. Say you spent five hundred dollars and Google Ads reported two thousand dollars in conversion value. That is 2,000 ÷ 500 = 4× return on ad spend on paper. But if GA4 only credits paid search with thirteen hundred dollars of that value, the same campaign reads as 1,300 ÷ 500 = 2.6×. Same spend, same real orders, two very different verdicts — and if you pause the campaign based on the lower one while it is genuinely profitable, the mistake costs you real margin. Deciding which figure to act on is the whole point of working out which is right, Google Ads or GA4 data.

Which number should you trust?

Neither, exactly — because they measure different things. Use each for its own job.

Trust Google Ads for questions about ad influence and for feeding the bidding algorithm; it is designed to answer "did my ad plausibly drive this?" Trust GA4 for cross-channel context and for seeing how paid search fits into the whole journey. And trust your Shopify order count for the one question both tools blur: how many sales actually happened and how much revenue landed.

The trap is comparing platform-reported ad numbers to each other and expecting a match. They are answering different questions in different currencies. The same trap catches sellers whose social pixels drift too, which is why Facebook ads not tracking Shopify purchases is the mirror image of this problem.

Where profit comes in

Here is the part every reconciliation article skips: none of these numbers is profit. ROAS on reported conversion value ignores product cost, fulfillment, processing fees, and refunds. A campaign at 4× ROAS can still lose money once a print-on-demand base cost and shipping come out.

That is where PodVector fits. It connects your Shopify, Meta Ads, Google Ads, Printify, Printful, and Stripe accounts and computes true per-order profit — the real margin left after ad spend and costs, not a platform's self-reported conversion value. Victor, its AI operator, reads that live data, flags where your reported ROAS and your actual profit diverge, and proposes moves you approve, executing them on the Shopify side. Victor does not touch your ad account, and PodVector is not a dashboard — it is an operator working from reconciled numbers. You can connect your stack and see true per-order profit to stop steering by whichever tool happened to claim the sale.

FAQs

Why does Google Ads show more conversions than GA4?

Usually three things stack up. Google Ads recovers conversions GA4 loses to ad blockers and consent declines, it credits itself under last-click while GA4 splits credit fractionally, and it layers extra modeling on top. Together those push the Ads number higher on most stores.

Is a difference between Google Ads and GA4 a tracking problem?

Not by default. A gap of roughly ten to thirty percent is expected from the structural differences in how the two count, according to Dataslayer. A gap far larger than that — or one that appears suddenly — is worth investigating for a broken tag, a consent-mode change, or a counting-setting flip.

How do I compare the two fairly?

Match the settings before you match the numbers. Use the same date range on a trailing window of a week or two rather than a single day, since Google Ads reports on the click date and GA4 on the conversion date. Confirm the time zones on the Google Ads account and the GA4 property agree, and check whether your conversion action counts "every" or "one per click."

Should I trust Google Ads or GA4 for reporting?

Use Google Ads to judge ad performance and feed bidding, and GA4 for cross-channel journey context. For the count of sales that truly happened, trust your Shopify orders. For the number that actually matters — profit after ad spend and costs — reconcile all of them rather than picking a favorite.

Will fixing my GA4 tags make the numbers match?

It will narrow the tracking-loss gap, but it will not close the methodology gap. Click-date versus conversion-date reporting, last-click versus data-driven attribution, and modeling remain even with flawless tags. Aim for a stable, explainable ratio between the tools, not an exact match.