GA4 and Google Ads conversions don't match, here's why
Why GA4 and Google Ads report different conversion numbers, attribution, counting, windows, consent and modelling, which one to bid on, and a reconciliation checklist.
GA4 and Google Ads will never report the same conversion numbers, they measure with different attribution scopes, counting methods, conversion windows, and modelling. A 15–30% gap is normal, not broken. This guide explains each source of difference, which number Smart Bidding should be fed, and a checklist for deciding when a gap is routine and when it's a real tracking fault.
Two systems answering different questions
The panic usually starts the same way: Google Ads says 214 conversions this month, GA4 says 163, and someone asks which one is lying. Neither. They're built to answer different questions:
- Google Ads asks: "how many conversions did my ads cause?" It's an advertising system giving itself credit for outcomes it touched.
- GA4 asks: "how did all channels together produce conversions?" It's an analytics system splitting credit across the whole journey: email, organic, social, direct, and ads.
A shopper who clicks a Shopping ad on Monday, returns via your newsletter on Thursday, and buys: Google Ads records 1 conversion (its click was in the path); GA4's cross-channel view gives that ad a fraction of the credit and the email the rest. Multiply across a month of traffic and the totals must diverge. A 15–30% gap is the normal operating state.
Knowing that, the useful work is understanding each specific source of difference: because a couple of them are configuration choices you can align, and one of them tells you which number to bid on.
The five sources of mismatch
1. Attribution scope and model
Google Ads attributes within its own walls: if an ad click appears in the path, Ads takes credit (data-driven attribution shares credit among Google ads touchpoints, not with your email channel). GA4's default reports attribute across all channels. This is the biggest single cause of Ads counting higher, and it's structural: see attribution for the fuller picture.
2. Counting method and timing
Google Ads reports conversions against the date of the click, GA4 against the date of the conversion. A Black Friday click converting on Cyber Monday lands on different dates in each system: so any date-range comparison is misaligned at the edges. Ads can also count every purchase after a click ("Every" counting) where a GA4 comparison might be sessions-scoped; and GA4 conversions can take 24–48 hours to finalise while the Ads tag reports within hours.
3. Conversion windows
Google Ads' default click-through window is 30 days (configurable 1–90). If your GA4 comparison or its acquisition logic effectively looks at a shorter path, long-consideration purchases fall out of one count but not the other. High-ticket products with 2–3 week decision cycles feel this hardest.
4. Consent and modelling
Since Consent Mode v2, both systems model conversions for users who decline tracking: but they model differently, from different signals, at different aggregation levels. In markets with high consent-banner rejection rates, a meaningful slice of both numbers is statistical estimate, and the estimates don't reconcile by design. Enhanced conversions recover some of this signal on the Ads side: which widens the gap against GA4 while making the Ads number more accurate, not less.
5. Tag and import plumbing
Everything above is methodology. This one is faults: duplicate tags firing twice, the purchase event missing on some templates, importing GA4 conversions into Ads alongside the native Ads tag (double-counting or double-discrepancy), currency mismatches, or a tag manager change that silently broke one system. Methodology gaps are stable; plumbing faults show up as sudden changes in the gap.
Which number should Smart Bidding eat?
For Shopping, feed Smart Bidding the native Google Ads conversion tag, enhanced conversions on, revenue values verified: not imported GA4 conversions. Three reasons:
- Speed. The Ads tag reports in hours; GA4 import adds up to a day or two of delay. Smart Bidding recalibrates on fresh signal: stale signal means sluggish bidding.
- Credit. GA4's cross-channel attribution hands part of every conversion to other channels before Ads sees it. Feed that to tROAS and Google under-values its own auctions, bids lower, and volume quietly shrinks: you've built a permanent handbrake into your bidding.
- Modelling fit. The Ads-side consent modelling is built to feed bidding; GA4's is built for reporting.
Keep GA4 as your honest cross-channel accountant. Bid on the Ads tag. Never run both as active conversion actions for purchases: one primary, the other secondary (observation only). Setup details in conversion tracking for Shopping.
Deep dive The reconciliation checklist
Work this in order when the gap worries you. The goal is not a perfect match: it's confidence that both systems are healthy and the difference is explained.
1. Establish ground truth first. Pull the month's actual order count and revenue from your store platform. Both GA4 and Ads should sit below it (no tracking captures everything). If either system reports more purchases than the store took, you have double-counting: usually a duplicate tag or a GA4 import running alongside the native tag. Fix that before comparing anything else.
2. Align the comparison before judging it. Same date range, plus a 3–7 day cooling-off buffer at the end (both systems restate recent days). Remember Ads reports by click date: compare a completed period, ideally 30+ days old at the edges. Compare purchase conversions only, not all conversion actions against all GA4 key events.
3. Size the gap. Ads 15–30% above GA4 for ecommerce: normal, stop here unless something changed. Under ~10%: unusually clean, enjoy it. Over ~40%, or a gap that suddenly moved: keep going.
4. Check the counting settings. In Ads: conversion action set to "Every" (right for purchases), window length, primary vs secondary status, and whether a GA4-imported action is also primary. In GA4: purchase event firing once per transaction (check a test order), and the attribution settings applied to the report you're reading.
5. Check the plumbing on both sides. Fire a test purchase and watch it arrive in both systems (Tag Assistant / GA4 DebugView). Confirm transaction IDs are sent: they're the deduplication key. Check the tag survives on all templates: the classic silent fault is a checkout or payment-provider page where the tag never fires, which suppresses both systems but rarely equally.
6. Check consent's share. High cookie-banner rejection in your market means more modelled conversions and a floatier gap. If you haven't set up enhanced conversions, that's the concrete improvement available on the Ads side.
7. Log the explained gap and monitor the change, not the level. Once you know your normal is, say, "Ads runs 22% above GA4", the useful alert is that ratio moving: a jump to 45% or a collapse to 5% means something broke this week. A stable gap is a healthy gap.
The mindset shift
Merchants lose real time trying to force these numbers to agree, and the effort is misdirected: the systems disagree because they're supposed to. The productive posture: verify both against store orders once, understand your normal gap, bid on the Ads tag, report cross-channel from GA4, and alert on changes in the relationship rather than the existence of one. A mismatch is a fact of measurement; only a moving mismatch is a fault.
Frequently asked questions
Why does Google Ads show more conversions than GA4?
Mainly attribution scope. Google Ads credits itself for any conversion its ads touched (including view-through on some campaign types), while GA4 splits credit across every channel in the path. Add Google Ads' longer default click window and by-date-of-click reporting, and Ads counting higher is the normal state, not an error.
How much difference between GA4 and Google Ads conversions is normal?
Gaps of 15–30% are routine for ecommerce, and larger during long-consideration purchase cycles. Investigate when the gap suddenly changes size, when one side flatlines, or when revenue (not just count) diverges sharply, those point to a tag or configuration fault rather than methodology.
Should I bid on GA4 conversions or Google Ads conversions?
For Shopping, bid on the native Google Ads tag (ideally with enhanced conversions) and keep GA4 as a cross-channel view. The Ads tag reports faster, supports modelling designed for bidding, and avoids GA4's added delays and cross-channel attribution which starve Smart Bidding of credit and speed.
Can I make GA4 and Google Ads match exactly?
No, and chasing an exact match wastes effort. The two systems answer different questions by design. What you can do is align attribution models and windows to narrow the gap, and verify both against your store's order count as the ground truth.