The future looks bright for Looker Studio, but there have been some consequences of the migration regarding its GA4 Data API quota. Previously unenforced, all users of Looker Studio and the GA4 Data API are now subject to the quota as detailed here.
Quota in Looker Studio
For Enterprise users of GA4 and Looker Studio, a common issue is hitting the limit on the number of concurrent API requests, which by default only allows ten at a time. Since each plot or table in a Looker Studio report may represent an API request, this could be exhausted with only one dashboard. Multiply that across an organisation that may have 10-100s of users looking at the dashboard, and quotas will be quickly exhausted. The impact is compounded by the fact that it is only the user who created the Looker dataset that has their quota used up, meaning one account may be servicing hundreds of others.
Quota increases are possible and may be an option, but they will only be up to a hard limit that still may be inadequate for your Looker Studio needs. They may also involve a lengthy application process.
The long-term solution is to use the GA4 data export to BigQuery, and connect to that instead. GA4 offers the BigQuery export to all, not just paying users, although there will be charges for the queries you make.
To help optimise the migration in costs and efficiencies, Devoteam’s data team can help in the process:
- temporarily increasing your team to help migrate existing reports & putting in place FinOps to monitor your usage across the organisation
- helping determining rules for access to various BigQuery datasets
- taking the opportunity to evolve your whole data stack to a fully fledged Google Cloud Platform system
A full data suite on Google Cloud can include adding other data sources such as CRM databases, providing data governance, separating critical business logic from within tools into its own dedicated role and creating data activation products beyond dashboards including your own APIs and utilising Google’s machine learning platform, Vertex AI.