Google Ads Conversions Not Matching GA4: Why

The symptom

Someone pulls up Google Ads’ conversions column next to GA4’s purchase count for the same date range, and they don’t match. Sometimes the gap is small and ignorable. Sometimes it’s large enough that budget decisions start hinging on which number someone happened to trust that day.

Direct answer

Google Ads and GA4 aren’t built to report identical conversion counts, even when both are correctly configured. Different attribution windows, conversion-time-versus-click-time accounting, and two entirely separate matching pipelines produce some baseline gap. The diagnostic question isn’t whether the numbers match exactly — structurally, they won’t — it’s whether the gap is stable and explainable, or growing and erratic. A gap that shrinks once the attribution window fully closes is structural and not a problem. A gap driven by lost gclids or an incomplete Enhanced Conversions setup is real, and it’s costing you attributed conversions you actually earned.

Causes, ranked

1. Attribution window mismatch. Google Ads’ conversion window (commonly 30–90 days depending on the action) counts a purchase on the date of the click, even weeks later. GA4 counts it on the date of the transaction. Compare a date range before that window has fully closed, and the two will disagree by design.
2. Click-through vs. view-through counting. Google Ads can count view-through conversions — an impression with no click that later converts — which have no equivalent session in GA4 at all.
3. Conversion time vs. click time. Related to the window issue but worth checking separately: switching a conversion action’s reporting between “by conversion time” and “by conversion date” can shift totals for the same underlying data.
4. gclid loss. The same redirect, consent, or cross-domain issues that cause GA4’s Unassigned traffic can also strip the gclid parameter before it reaches Google Ads’ matching logic — meaning a real ad-driven purchase never gets credited to the ad at all.
5. Enhanced Conversions gaps. If hashed customer data isn’t being sent, or is incomplete, conversions that could have matched via Enhanced Conversions fall back to weaker cookie-based matching, or don’t match at all.
6. Structural counting-model differences. Even with everything else correct, GA4 and Google Ads use separate attribution models to credit conversions across their own data. Identical underlying orders can still produce two different, both-legitimate counts.

How to diagnose it

1. In Google Ads → Goals → Conversions → Summary, open the specific conversion action and check its attribution model and status.
2. Check the conversion window settings on that action — click-through and view-through windows are set independently and both affect the count.
3. Compare totals only for a date range where the attribution window has already closed (for example, a 60–90 day old month) rather than the trailing 7 or 14 days, where the numbers are still settling.
4. In the conversion action’s settings, check whether it’s reporting by conversion time or by conversion date, and note how much the total shifts between the two views.
5. Confirm auto-tagging is on (Google Ads → Account settings), then manually click through a real ad-tagged URL to checkout to verify the gclid parameter survives every redirect.
6. In Google Ads → Tools → Conversions, check the Enhanced Conversions status column for a “needs attention” flag or a visible match-rate indicator, and confirm hashed customer data is actually present in the network request on the purchase page.
7. In GA4 → Advertising → Attribution → Model comparison, compare GA4’s own conversion counts for the Google Ads channel under different models, to see how much the choice of model alone moves the number independent of any tracking issue.

Decision table

Cause	Signal you’ll see	Fix
Attribution window mismatch	The gap shrinks the further back you look, and converges for fully closed date ranges	No fix needed — compare only closed windows, and document the expected lag for stakeholders
View-through inflation	Google Ads includes conversions with no matching click or session in GA4	Check the view-through window setting; decide deliberately whether to include it in comparisons
gclid loss	Click-based conversions run consistently below actual paid orders, tracing to a redirect or consent path	Fix the parameter-stripping redirect; confirm auto-tagging is on and gclid persists through checkout
Enhanced Conversions gap	Match rate is low or partial in the Ads UI; hashed data missing from the network payload	Complete the Enhanced Conversions data flow (email, phone, address) on the purchase page
Structural counting-model difference	The gap is stable and roughly consistent month to month, and doesn’t trend toward zero	Document it as expected; treat stability, not size, as the health signal

Fix it yourself, or bring in help

Checking attribution window and conversion-timing settings, confirming auto-tagging is on, and comparing only fully closed date ranges is something you can do directly in the Google Ads and GA4 interfaces. What usually needs a specialist is implementing or debugging Enhanced Conversions’ data flow, tracing gclid loss through a redirect or consent chain you don’t fully control, or building a repeatable reconciliation process across Ads, GA4, and Shopify rather than eyeballing two dashboards each month. If the gap is growing rather than converging, or shows up right after a checkout or consent change, that’s worth scoping through Google Ads conversion tracking work rather than assuming it’s just structural noise.

Related reading

Redirects and consent issues that strip gclid overlap heavily with what causes GA4 unassigned traffic — the same diagnostic steps apply to both. For the broader question of how GA4 revenue and Shopify’s own numbers relate, see GA4 revenue doesn’t match Shopify.