Why don't Google Ads and GA4 conversions match?
Google Ads and GA4 conversions don't match because the two products answer different questions. Google Ads asks which Google ad interaction led to a conversion, while GA4 spreads credit for each key event across every channel it saw, including organic search and other ad platforms. Google's own help centre calls discrepancies "common", even "with a correct setup" (Google Ads Help). These are the seven settings that create the gap.
| Setting | Google Ads | GA4 |
|---|---|---|
| Date a conversion is reported on | Date of the ad click | Date the key event happened |
| Who can get credit | Google Ads interactions only | Paid and organic channels in GA4 reports; Google paid channels by default in what it shares with Google Ads |
| Attribution model | Data-driven by default for most conversion actions; last click selectable | Data-driven, marked recommended; paid and organic last click or Google paid channels last click selectable, as one property-wide setting |
| Default window | 30 days after a click (1 to 90 selectable); 3 days after an engaged view; 1 day after an impression | 90 days (30 or 60 selectable); 30 days for first_visit and first_open (or 7) |
| Counting | Every or One per click, set per conversion action | Once per event (default) or once per session |
| Modeled conversions | Included in the Conversions column | Included when consent mode modeling applies |
| Time zone | Ads account time zone | GA4 property time zone |
Sources: Google Ads Help and Analytics Help pages linked in each section below, checked October 2026.
1. Click date versus event date
Google Ads reports a conversion on "the date and time of the click that led to the successful action", while GA4 reports it on the day it occurred (Google Ads Help). A click on 28 September that converts on 3 October is a September conversion in Google Ads and an October key event in GA4. Monthly totals straddling that boundary disagree. The most recent days in Google Ads also keep growing for weeks as late conversions arrive. Google Ads has columns that use the other basis, and for reconciling with Google Analytics, Google suggests "All conv. (by conv. time)" (Google Ads Help). Use the by-conversion-time columns whenever you compare with GA4, and give the last 48 hours time to settle.
2. Attribution model and who can get credit
Data-driven attribution is the default for most Google Ads conversion actions (Google Ads Help). GA4's reporting attribution model is one setting for the whole property, made in Admin, and changing it rewrites historical data as well as future data. The options are data-driven or last click across paid and organic channels, or last click across Google paid channels, which falls back to paid and organic last click when there was no Google Ads click (Analytics Help).
The bigger effect is the pool of candidates. In Google Ads, only Google ads compete for the credit. In a GA4 report on paid and organic channels, an organic search visit, a newsletter click or a Meta ad can take some or all of it. GA4's own help says its totals "may be higher than what you find in Google Ads", because Google Ads includes only conversions attributed to Google paid channels (Analytics Help). Data-driven attribution in GA4 can reattribute a key event for up to 7 days after it happens (Analytics Help), and attributed key event data can update for up to 12 days (Analytics Help), so last week's GA4 figure may still move.
3. Conversion windows
A Google Ads conversion action counts conversions up to 30 days after a click by default, 3 days after an engaged view and 1 day after an impression; the click window can be set anywhere from 1 to 90 days (Google Ads Help). GA4 looks back 90 days for most key events by default, with 30 or 60 days selectable, and 30 days for first_visit and first_open (Analytics Help). On a product with a long consideration cycle, at default settings, a click 45 days before purchase counts in GA4 and not in Google Ads.
4. Counting every conversion or one
Google Ads counts either every conversion after a click or only one per click (Google Ads Help). GA4 counts a key event every time it fires by default, or once per session for key events created from old Universal Analytics goals (Analytics Help). A lead form set to "One" in Google Ads and "once per event" in GA4 diverges whenever someone submits twice.
5. Imported GA4 key events versus the Google Ads tag
You can measure the same purchase twice: once with the Google Ads tag and once as a GA4 key event imported into Google Ads. Imported GA4 data reaches Google Ads up to 24 hours after the event (Google Ads Help), view-through conversions are not available on imported conversions, and cross-device conversions need Google signals switched on in the GA4 property (Analytics Help). If both versions are set as primary, the Conversions column counts the purchase twice, because Conversions sums every primary action (Google Ads Help).
6. Consent mode, modeled conversions and Enhanced Conversions
This cause weighs most in Europe, where GDPR consent banners let visitors decline tracking. Google Ads models conversions for unconsented users, and its Conversions column shows "both modeled and observed conversions" (Google Ads Help).
GA4 runs two separate systems, and they are easy to mix up:
- Key event modeling fills in key events from unconsented users when you use consent mode. Google publishes no volume threshold for it, only that "if there isn't enough traffic to inform the model", modeled key events aren't reported, and in GA4 those key events are attributed to Direct (Analytics Help).
- Behavioural modeling estimates users and sessions, and Google says it "is not applied to event counts". It needs advanced consent mode, with tags loading before the banner, plus at least 1,000 events a day with analytics_storage denied for at least 7 days and at least 1,000 daily users with it granted on at least 7 of the previous 28 days (Analytics Help).
So the behavioural thresholds decide whether unconsented users and sessions are modeled in GA4, and say nothing about whether its key event totals match Google Ads. A mid-sized shop in the Netherlands or Austria with too little traffic for key event modeling sees Google Ads model the missing conversions while GA4 attributes the unconsented key events to Direct, which also shrinks the google / cpc count you compare against.
Enhanced Conversions sends a SHA-256 hash of the email or phone number a customer entered, so Google can match the conversion to a signed-in Google account (Google Ads Help). GA4 has its own route to the same data, user-provided data collection (Analytics Help). Turning on one and not the other shifts the gap gradually: Google says the effect on models, attribution and bidding "may take up to one month to appear".
7. Cross-device journeys and time zones
Google Ads credits a click on a phone that converts on a laptop, reports it in a separate "Cross-device conv." column and includes it in Conversions. GA4 joins the two sessions only through a User-ID you send, Google signals for signed-in users, or its modeling. Time zones are simpler: GA4 reports in the property's time zone and Google Ads in the account's, and Google warns that different settings cause discrepancies (Analytics Help). An account on UTC and a property on Berlin time shift every daily total by one or two hours.
How do you reconcile Google Ads and GA4 conversions?
Work through these steps in order. The first four remove differences in how you compare. The rest look for broken tracking.
- Match time zones. Check the Google Ads account and the GA4 property. If they differ, compare weekly totals only.
- Compare on the same date basis. Use "All conv. (by conv. time)" in Google Ads, or switch GA4's conversion performance report to interaction time.
- Compare the same scope. In GA4, filter to session source / medium google / cpc, or open Advertising, Conversion performance and view it for Google Ads (Analytics Help). Leave the property's attribution model alone for this: it is a property-wide setting that rewrites historical data, and the Google paid channels model still falls back to paid and organic last click when there was no Google Ads click.
- Match counting and windows. Set each pair of matching actions to the same counting method and as close a window as each tool allows.
- Look for double counting. In Google Ads, list the primary conversion actions. A Google Ads tag purchase and an imported GA4 purchase should never both be primary.
- Check auto-tagging survives. Click a live ad and confirm the gclid parameter reaches the landing page. Redirects, consent tools and URL rewrites can strip it, and GA4 then files the visit under organic or direct.
- Check consent and Enhanced Conversions on both sides. Confirm consent mode sends all four parameters, including ad_user_data and ad_personalization, and that Enhanced Conversions and GA4 user-provided data collection are either both on or both off.
- Check against your backend. Pull orders or leads from Shopify, your CRM or your database for the same week. Whichever tool drifts away from the backend is the one with the problem.
How big a gap between Google Ads and GA4 is normal?
Google's only published figure is in Google Ads Help: "Discrepancies, often up to 20%, are expected due to different attribution models" (Google Ads Help). We found no sourced benchmark that holds across accounts, so the band to track is your own. Your gap depends on your consent rate, sales cycle, channel mix and the settings in the table above, so an account that has run at a steady 25% for a year is healthier than one that jumped from 5% to 15% last week.
What you can measure is whether your own gap is stable. Each week, divide Google Ads conversions (by conversion time) by GA4 key events from google / cpc. Then set a band at the eight-week mean, plus or minus two standard deviations.
| Week | Google Ads conv. | GA4, google / cpc | Ratio |
|---|---|---|---|
| 1 | 1,180 | 1,000 | 1.18 |
| 2 | 1,215 | 1,004 | 1.21 |
| 3 | 1,162 | 1,002 | 1.16 |
| 4 | 1,240 | 1,033 | 1.20 |
| 5 | 1,198 | 1,007 | 1.19 |
| 6 | 1,226 | 1,005 | 1.22 |
| 7 | 1,171 | 1,001 | 1.17 |
| 8 | 1,209 | 1,016 | 1.19 |
| 9 | 1,236 | 922 | 1.34 |
Example figures for a German online shop.
Say a German online shop logs the nine weeks above. Over weeks 1 to 8 the ratio averages 1.19 with a standard deviation of 0.02, so the band runs from 1.15 to 1.23. Week 9 comes in at 1.34. Google Ads conversions look normal, while GA4 is about 87 key events below its eight-week average of 1,008.5. That pattern points at the GA4 side, typically a consent banner update or a checkout redirect that strips the gclid. A ratio that drops below its band points the other way, usually at the Google Ads tag or an import that stopped.
Recompute the band after any change you make on purpose, such as switching on Enhanced Conversions, so the new normal doesn't look like an incident.
Which number should you use for which decision?
| Decision | Use | Why |
|---|---|---|
| Bids and targets inside Google Ads | Google Ads conversions | Smart Bidding optimises on the primary conversion actions, so targets must use the same numbers |
| Splitting budget between Google, Meta and other channels | GA4 or backend numbers, corrected by incrementality tests | Each platform credits itself; you need one scale across channels |
| Revenue and CAC reported to finance | Backend orders or CRM revenue | It's the number the business books, before any attribution |
| Spotting tracking breaks | The weekly ratio between the two | A jump outside the eight-week band flags a tracking incident on one side within a week |
For the question of whether Google Ads conversions are incremental at all, attribution in either tool can't answer it. That takes a test: see incrementality testing and the geo lift test guide. If Meta is part of your mix, its numbers disagree with GA4 for a different set of reasons, covered in why Meta Ads conversions don't match GA4.
Both numbers side by side, every week.
Nupact shows platform-reported and reconciled conversions next to each other for Google Ads, Meta, TikTok and Amazon Ads, and the gap stays visible in every report. Each platform's own figures are shown against your GA4 and backend data. A gap that moves outside its usual band reaches you as an alert in Slack, Teams or email.
On the Google side, enriched conversions go back through Enhanced Conversions, and Nupact monitors the match rate and runs a dry run before the first upload.
What else should you know about Google Ads and GA4 counts?
Why does Google Ads show more conversions than GA4?
Usually because Google Ads includes modeled and cross-device conversions that GA4 can't see, and keeps all the credit for Google ads while GA4 shares it with other channels. Check also that a Google Ads tag action and an imported GA4 action aren't both set as primary, which counts each conversion twice.
Why does GA4 show more conversions than Google Ads?
GA4 counts key events from every source, so filter it to google / cpc before comparing. Within Google traffic, GA4's 90-day lookback can also catch conversions after a Google Ads window of 30 days or less has closed.
Should I import GA4 conversions or use the Google Ads tag?
For bidding, the Google Ads tag with Enhanced Conversions is usually the better source: it reports without the up-to-24-hour import delay and supports view-through and cross-device data. Imported GA4 key events suit actions only GA4 measures. Whichever you pick, set only one version of each conversion as primary.
Does consent mode make Google Ads and GA4 match?
No. With consent mode, Google Ads models conversions from visitors who decline cookies, and GA4 models their key events only when there is enough traffic to inform the model; otherwise GA4 attributes those key events to Direct. GA4's published thresholds (1,000 denied events a day for at least 7 days, 1,000 granted daily users on at least 7 of 28 days, advanced consent mode) govern behavioural modeling of users and sessions, which doesn't touch event counts. In Europe consent mode tends to change the gap, so reset your baseline band after implementing it.
Bring one Google Ads account and its GA4 property.
We put platform-reported and reconciled conversions side by side for your last eight weeks and show where the gap moved. Founder-led, your setup on screen.