Linking Google Ads to GA4 is presented as a single click in the admin interface, but "linked" doesn't mean "everything syncs automatically." Different data types travel in different directions, some require extra configuration, and a few things that practitioners expect to just work quietly don't.
This guide is for PPC managers and analysts who have already linked — or are about to link — and want a clear picture of what they're actually getting.
Why the link matters at all
Before the link, Google Ads and GA4 operate as isolated silos. Ads knows what it spent and what clicks it generated; GA4 knows what users did after they landed. The link creates a data bridge in both directions: GA4 starts receiving cost and campaign data from Ads, and you gain the ability to push GA4 conversion events and audience lists back into Ads for bidding and targeting.
Without the link, you're flying with two separate instruments that don't talk to each other. With it, you can close the loop between spend and on-site behaviour — but only for the data types that actually flow across it.
The data flow matrix
Here's what actually moves, in which direction, and what caveats apply.
| Data type | Direction | Requires link? | Notes & caveats |
|---|---|---|---|
| Auto-tagging (gclid) | Ads → GA4 | No — but link is needed for cost data to follow | gclid is appended to every Ads click URL automatically. GA4 reads it to populate campaign dimensions. Auto-tagging must be enabled in your Ads account settings. |
| Cost data | Ads → GA4 | Yes | Clicks, impressions, cost, and ROAS appear in GA4's Advertising reports. Data can lag 24–48 hours. Cost data will not appear if auto-tagging is disabled or gclid is stripped by redirects. |
| Conversion imports | GA4 → Ads | Yes | GA4 key events can be imported as Ads conversion actions. You choose which key events to import; they don't all flow automatically. Imported conversions count in Ads reporting and can be used for Smart Bidding. |
| Audience segments | GA4 → Ads (bidirectional for use) | Yes | GA4 audiences can be published to Ads for remarketing and Customer Match. Audiences must meet a minimum size threshold (typically 1,000 users for Display, 100 for Search). GDPR/consent restrictions apply. |
| Campaign performance | Ads → GA4 | Yes | Campaign name, ad group, keyword, match type, and network appear as dimensions in GA4. Available in Explorations and the Advertising snapshot. Not available in standard Acquisition reports without customisation. |
| Click attribution | Ads → GA4 (via gclid) | No (but link improves fidelity) | Attribution is handled independently in each platform. GA4 uses its own attribution model (data-driven by default); Ads uses its own. Numbers will differ — this is expected, not a bug. See the note on attribution models below. |
Auto-tagging: the foundation everything else depends on
Auto-tagging is Google Ads appending a gclid parameter to every click URL. This unique identifier is how GA4 knows a session came from Google Ads — not just from the generic google / cpc source/medium, but from the specific campaign, ad group, and keyword that drove the click.
Without auto-tagging, GA4 can only see that traffic came from Google organic or a manually tagged URL. With it, GA4 can pull full campaign-level detail from the click. Auto-tagging is enabled by default in new Ads accounts, but it can be accidentally disabled — check under Google Ads Account Settings > Auto-tagging if your campaign data is missing in GA4.
Using UTMs and auto-tagging together? Auto-tagging wins for Ads attribution — the gclid takes precedence over any UTM parameters for session attribution back to Google Ads. UTMs still appear in GA4 as dimensions, so your utm_campaign values are visible and reportable. They just don't override the gclid for the purpose of attributing the session to Ads.
This means if you're manually tagging Ads URLs with UTMs for a specific reason (custom campaign names, cross-account tracking), those UTM values will show in GA4, but Ads attribution will use the gclid. Plan for both. See UTM parameters vs auto-tagging for a fuller comparison.
Importing GA4 conversions into Google Ads
This is the most commercially important flow and the one most often configured incorrectly. When you import a GA4 key event as an Ads conversion action, you're telling Smart Bidding: "optimise for this." What you choose here directly affects how your campaigns spend.
A few things practitioners frequently get wrong:
- Importing micro-conversions alongside macro-conversions dilutes bidding signals. If you import "scroll 50%" and "purchase" as equal conversion actions, the algorithm treats a scroll as equivalent to a sale. Set micro-conversions to "observation" rather than "primary" — or don't import them at all.
- Duplicate counting can occur if you have both a GA4-imported conversion and a natively tracked Ads conversion (e.g., from a Google Ads tag) measuring the same action. Audit your Ads conversion actions and ensure each action has one source.
- The conversion window in Ads is separate from GA4's lookback window. A 90-day conversion window in Ads will count conversions that GA4's standard reports may not, especially if you've altered GA4's attribution lookback. The numbers will legitimately differ — this is normal.
For a deeper look at how GA4 designates which events become conversion candidates, see Conversions vs Key Events in GA4.
Audience sharing for remarketing
Once linked, GA4 audiences become available in Google Ads for remarketing campaigns and Customer Match. You build the audience in GA4 — based on events, user properties, or predictive metrics — and publish it to Ads, where it appears as a targeting segment.
This is one of the most powerful capabilities the link unlocks, particularly for:
- Excluding recent converters from acquisition campaigns
- Targeting users who abandoned checkout
- Building lookalike segments from high-value purchasers
- Suppressing known customers from brand awareness spend
The catch is consent and size thresholds. Under GDPR and similar frameworks, users must have consented to personalised advertising. GA4's consent mode integration handles some of this, but you should verify your consent setup before relying on audience targeting in restricted markets. Audience populations below Google's minimums will show as "Too small to target" in Ads and won't serve.
The "cost data not showing" problem
This is the most common issue practitioners hit after completing the link. The account is linked, but the Advertising reports in GA4 show no cost data, or cost appears for some campaigns and not others. The usual causes:
- Auto-tagging is disabled. Re-enable it in Ads account settings. Historical gaps won't backfill.
- A redirect strips the gclid. If your landing page URL goes through a redirect (e.g., a tracking template, a campaign management platform, or a CMS redirect) that doesn't preserve URL parameters, the gclid is lost. Check your redirect chain with a test click and inspect the final URL for the gclid parameter.
- The link was made with insufficient permissions. The Google account that created the link needs at least "Editor" access in GA4 and at least "Standard" access in Google Ads. Links created with read-only access may appear to succeed but won't activate cost data import.
- Data lag. Cost data in GA4 can be 24–48 hours behind. If everything looks correct but no data has appeared, wait two days before investigating further.
- Multiple Ads accounts, partial link. If a single GA4 property receives traffic from more than one Ads account (common in agency setups), each account needs its own link. Traffic from unlinked accounts will show campaign data via gclid but not cost data.
For client-facing Ads reporting that draws on linked data, see building client-ready Ads reports.
How to link the accounts correctly
The official path is through GA4 admin, not Google Ads admin — it gives GA4 the authority over what gets linked.
- In GA4, go to Admin > Property > Product Links > Google Ads Links and click Link.
- Select the Google Ads account(s) you want to connect. You'll see a list of accounts your Google login has access to. If an account is missing, confirm you have at least Standard access in that Ads account.
- Enable Personalised advertising if you want to use GA4 audiences for remarketing. You can leave this off and still get cost data and conversion imports — it only gates the audience sharing feature.
- Enable Auto-tagging if prompted. If the link wizard detects auto-tagging is off, it will offer to turn it on — accept this unless you have a specific reason not to.
- Review and confirm. The link activates within a few hours. To import conversions, go separately to Google Ads > Tools > Conversions > Import > Google Analytics 4 properties and select the key events you want to import as conversion actions.
Why the numbers still won't perfectly match
Even with a correctly configured link, clicks in Ads will not equal sessions in GA4, and conversions in Ads will not equal conversions in GA4. This is not a sign something is broken — it is the product of how each platform counts.
Ads counts a click when someone clicks the ad. GA4 counts a session when a session_start event fires on your site. Between the click and the session start, users may bounce before the tag loads, be blocked by ad blockers, navigate away during a redirect, or land on a page where GA4 is not implemented. Industry-standard click-to-session loss is typically 5–15%.
On conversions, the two platforms use different attribution models with different lookback windows and different de-duplication logic. Some divergence is structural and permanent. The goal is not to make the numbers identical — it's to understand why they differ and which number to use for which decision.
Summary
The Google Ads–GA4 link is genuinely useful, but it's a bridge with lane markings, not a fire hose. Cost data and campaign dimensions flow from Ads to GA4; conversions and audiences flow the other way. Auto-tagging is the foundation — without it, most of the value disappears. And no amount of correct configuration will make Ads and GA4 report identical numbers, because they're measuring adjacent but distinct things.
Get the link right once, import the right conversions, and the two platforms become meaningfully complementary. Leave it misconfigured and you'll be reconciling spreadsheets forever.