How to link GA4 and Google Ads - Zephyra Studio

Linking GA4 and Google Ads makes it possible to see in analytics how much was spent on ads and what that traffic did on the site, in one report. The process has three steps: linking the accounts in the analytics settings, enabling auto-tagging in Ads, and deciding which conversions get imported. Linking requires administrative rights on both accounts, and the numbers do not join up on their own: measurement on the site also has to be set up correctly, because otherwise analytics receives the cost without the result. This article covers the steps, the requirements and the settings that change the numbers, including the most common mistake, which is counting the same conversion twice.

What linking makes possible

The first benefit is a shared report. Analytics shows how much was spent on ads, how many sessions arrived from them and how many of those visitors left an enquiry, so the channel is judged on the result rather than on click counts.

The second benefit is importing conversions into Ads. Events already measured cleanly in analytics can be used as conversions in advertising, which helps where the site has no separately configured tags for each action.

The third benefit is audiences. Groups of visitors can be built from analytics, for example people who viewed a service page and left without an enquiry, and those groups are used in ads for follow-up contact.

All three benefits depend on the same condition: that measurement on the site works before linking. If events are not set up, linking produces a prettier report about spending without an answer to what that money brought.

The linking steps

The process is short and takes about ten minutes, but the result can only be checked a day or two later, once the first data appears.

• Check that you have administrative rights on both accounts, because linking is not possible without them
• In analytics, open the product links section, choose Google Ads and link the right account
• In Ads, turn on auto-tagging, because it adds an identifier to each click that analytics can read
• Decide which events to import as conversions, because not all of them should be primary
• Wait a day or two and check that sessions from ads appear with the correct campaign names
• Check that conversions imported from analytics appear in Ads and are not counted twice

What changes in the reports after linking

After linking, the traffic reports include a section showing paid campaigns, with the amount spent alongside sessions and conversions. For the first time the relationship between what went in and what came out is visible in one report, without switching between two tools.

Channels can also be compared under the same measurement. When organic and paid traffic are read through the same events, it becomes clear which channel brings enquiries at a lower cost and which brings enquiries that turn into work more often.

Historical data from before the link stays without cost, so comparisons across it are not made casually. On accounts that ran for years without a link between the tools, the first month after linking establishes a baseline rather than a verdict.

Auto-tagging and campaign tags

Auto-tagging adds an identifier to every click, which analytics uses to connect a visit with the exact campaign, ad group and keyword. Without it, data from advertising cannot be connected to sessions, so the traffic stays in the general total.

Manual campaign tags are used where a link does not run through Ads, for example in e-mail or on social media. For ads it matters that the source is google. In paid campaigns the medium has to be tagged as cpc, because otherwise the traffic appears as a referral or as part of organic, and the advertising report stays empty.

When both auto-tagging and manual tags are used, they must not be mixed in the same link without reason. The most common mistake is adding manual tags to ad links, which overrides auto-tagging and loses the keyword data.

If the site runs on several domains or subdomains, measurement is configured to follow the journey between them, because otherwise the visit ends at the first domain and the path to the enquiry is broken.

What to check after linking

The first check is whether cost and sessions appear in the advertising reports. If the cost is there but sessions are not, auto-tagging is usually switched off, or the linking was done on the wrong account.

The second check is whether imported conversions arrive in Ads and whether they are marked primary or secondary. If every imported conversion becomes primary, the algorithm optimises towards actions that are not business.

The third check is whether the same action is measured twice, once by an on-site tag and once by the analytics import. The report then shows two enquiries for one submitted request, so cost per conversion is also calculated wrongly.

The fourth check is the relationship between the numbers in the two tools. Analytics and Ads attribute conversions differently and use different windows, so their figures do not match as a rule. Comparing makes sense over time within one tool, not between two.

Mistakes that spoil the data

All the common mistakes end the same way: the report looks tidy while the decision rests on a wrong number.

• Linking without auto-tagging, so cost exists without any data about the campaign
• Importing every event as a primary conversion, which dilutes the optimisation
• Measuring the same action both by an on-site tag and by the analytics import, counting it twice
• Adding manual tags to ad links and overriding auto-tagging
• Internal traffic and test visits not excluded, so enquiries from the office count as results
• Judging the outcome on day one, before the data has been processed at all

Source

Key takeaways

• Linking joins advertising spend with on-site behaviour, but it does not repair measurement that was never set up.
• Without auto-tagging, cost exists while data about the campaign and keyword does not.
• Import only a small number of events as conversions, because optimising for many goals optimises for nothing.
• The same action must not be measured by both an on-site tag and an import, because then it counts twice.
• Figures in analytics and Ads do not match as a rule, so compare over time within one tool.

Conclusion

Linking is a ten-minute job that changes how reports are read, because cost and result end up in the same place for the first time. If the numbers still look odd after linking, the cause is almost always the measurement on the site rather than the link itself. We set up tracking and link the accounts as part of our Google Ads service, and if you run the ads yourself it is worth reading our article on tracking conversions in Google Ads first, because it contains the checks that prevent double counting.