Integrations
Google Analytics 4
Connect GA4 through its BigQuery export to add sessions, conversions and channel attribution to a domain, and to see which entry pages bring the visitors who buy.
Search Console tells you who arrived from search. GA4 tells you what they did next. Connected to a domain, it adds sessions, engagement and conversions, and powers a set of attribution reports: which channels get credit for conversions, which campaigns pay back, and which landing pages bring the visitors who buy later.
AmICited reads GA4 through its BigQuery export, not the GA4 reporting API. The export carries raw, unsampled events, which is what makes the attribution reports possible. GA4 is available on every plan.
Before you start#
- A GA4 property with BigQuery export enabled. In GA4, go to Admin → Product links → BigQuery links and link the property to a Google Cloud project. The property cannot be selected in AmICited until the export exists. If you do not have a Cloud project, AmICited can host the export for you instead (Path B below).
- The right Google account. Use a Google account that can read the property. For the hosted export, that account needs the Editor role on the GA4 property.
- Workspace admin rights. Only workspace administrators can connect integrations.
Connect GA4#
Open Data Sources
Go to Data Sources and click Connect on Google Analytics 4 (under Organic).
Sign in with Google
Authorize with the Google account that has access to the property.
Select the property
Click Select property and choose the GA4 property for this domain. AmICited then checks that it can read the property’s export dataset (analytics_<propertyId>).
Choose history and confirm
Available history begins when your BigQuery export began; GA4 does not export events from before the link was created. Pick a window, review the projected rows and cost, and click Confirm and backfill. After that, new days import daily.
If AmICited cannot read the export#
The most common failure is a Google Cloud permission issue, not a GA4 one, so reconnecting GA4 does not help. The error shows the Cloud project number and the dataset name. There are two ways to fix it.
Path A: grant access
Someone with IAM admin rights on the export’s Google Cloud project does this:
- Open the Google Cloud Console and select the project by the number in the error (the project picker searches project numbers).
- Go to IAM & Admin → IAM → Grant access.
- Add the Google account that connected GA4 as a principal.
- Assign BigQuery Job User (
roles/bigquery.jobUser). AmICited runs its import queries as jobs in this project, so Data Viewer alone is not enough. - Assign BigQuery Data Viewer (
roles/bigquery.dataViewer), either for the project or only on the export dataset: in BigQuery, find analytics_<propertyId>, then Sharing → Permissions → Add principal. - Save, return to AmICited, and retry the connection.
Path B: hosted export
When the connection panel offers Host this export, AmICited creates the BigQuery link through GA4 Admin and stores the data itself. No Google Cloud project or IAM grant is needed; the connected account needs Editor on the GA4 property.
- The export is forward-only: it starts now and includes no history.
- Data appears within a day; until then the card shows that Google is still setting up the export.
- Your workspace retention setting applies to the hosted data.
- This option is not offered when the property already exports to another project.
What it brings and which reports use it#
| Report | Question it answers |
|---|
| Attribution | Which channels receive credit for conversions over time? Switch between first touch, last touch and linear models. |
| Conversions by location | Where do attributed conversions happen? |
| Campaign performance | Which campaigns return more than they cost? |
| Landing pages | Which pages get traffic and produce no outcome? |
| Acquisition revenue | Which entry page brought the visitors who bought later, and what their organic queries were worth? |
All five are under Reports → Platform reports → Google Analytics 4. Acquisition revenue is the one that ties back to AI visibility: pages that AI engines cite and that also bring buyers are the ones to protect first.
Settings#
- Attribution window. How many days before a conversion a touchpoint can still get credit (Lookback days).
- Cleansing rules. Rules applied before GA4 data is reported, such as how conversion value is taken (the event value or a fixed value), with diagnostics showing excluded and quarantined events for a date range.
- Identity spaces. Group domains that may share customer identities supplied by your application, for sites that share one customer base.
Cost and disconnecting#
GA4 follows the same model as every data source: a one-time import billed by rows kept (1 credit per million rows) and a flat daily warehouse fee per connected domain, set by your plan and shared with the domain’s other sources. Both are shown before you confirm. Disconnect stops the import and deletes the imported data; it cannot be undone. See Search Console and Bing for the full cost and sync-status details.
