Reports

[GA4] Data freshness and Service Level Agreement constraints

This article is for website and/or app owners who wish to know more about data freshness and the impact of the Analytics 360 data-processing and reporting Service Level Agreements (SLAs) on features.  

Data Freshness

Data freshness refers to how recently we've collected, processed, and reported data in your property. If processing takes 20 minutes, then the data freshness is 20 minutes.

Freshness varies because data is processed at different intervals. For example, intraday data is updated throughout the day for quicker access to some of your data, while daily data is more complete (drawing on more data sources) and available once per day for data with longer processing times.

We may take 24-48 hours to process data. During that time, data in your reports may change.

Intervals of data freshness

Analytics has the following typical intervals of data freshness for Google Analytics 360 and/or standard properties:

Interval Typical processing time Properties Data limits per property Query coverage
Realtime Typically less than 1 minute 360, Standard None Limited to a few dimensions and metrics
360 intraday About 1 hour 360 Premium Normal and Premium Large as defined here All reports and API queries, except these
Standard intraday 4-8 hours Standard Standard Normal All reports and API queries, except these
Daily 12 hours 360, Standard Standard, Premium Normal All reports and API queries
Daily 18 hours 360, Standard Premium Large All reports and API queries
Daily 24+ hours 360, Standard Premium XLarge All reports and API queries

About the intervals

Each day, we process data in three intervals—realtime, intraday, and daily—according to the processing times above. These intervals enable you to access some of your data more quickly as it becomes available.

Realtime

Realtime data is the most up-to-date data set and lets you monitor activity as it happens, but it covers fewer features than other intervals.

Intraday

Intraday data provides some data from the previous day, refreshing multiple times throughout the day so you can access data from the previous day more quickly; for 360 properties, intraday data is continuous. It's typically available before daily data in reports and API queries. When the intraday data is available before the daily data on any given day, you can expect the following:

  • You may notice temporary gaps in some event-scoped traffic source dimensions (such as source, medium, campaign, and default channel group). These temporary gaps can happen due to delays in receiving data for example from third-party conversion sources.
  • If event-scoped traffic source dimensions are available, we'll use the Paid and organic last click attribution model by default until the daily data is available.
  • We may apply stricter cardinality limits until daily data is available. As a result, you're more likely to experience the (other) row during this period if your property has high-cardinality dimensions.

Note: If intraday data is available before daily data, the data in reports and API queries may change after the daily data becomes available.

Daily

Daily data represents all the data for a day. If your report uses daily data, you can expect the following:

  • If event-scoped traffic source dimensions are available, we'll use the attribution model selected for the property.
  • Attribution credit for key events can change for up to 12 days after the key event is recorded, as Analytics' key event modeling improves.

Prior day data processing timeline

This table lists when data is typically available* in the property’s timezone for the prior day for most properties; actual processing times can vary.

Typical availability Data type
12:00 am Daily collection stops**
5:00 am 98% of BigQuery fresh daily export (beta) data is available for 360 customers
11:30 am Daily data is ready in Explore
12:00 pm Daily BigQuery events are ready
3:30 pm Daily data is ready in Reports

* This is the typical processing time that most data is usually available by. This is not a guarantee, nor an SLA or an SLO. Data is sometimes delayed beyond these processing times, particularly for large properties, complex data, or during uncommon processing slowdowns.

** Some data can arrive late, at times up to 7 days delayed 

Property categories

On any given day, a property is categorized based on the number of events it has collected and processed. A property is categorized as:

  • "Normal" if the property has collected and processed fewer than 25 billion events
  • "Large" if the property has collected and processed 25 billion or more events
  • "XLarge" if the property has collected and processed 250 billion or more events

These measurements are taken from the previous 31-day period, excluding the current day in the property's timezone.

Note that a property that is typically categorized as "Normal" or "Large" can be considered categorized as "XLarge" for a day if it has been collecting and processing an average of 15 billion or more events per day over the previous seven days, excluding the current day in the property's timezone.

Limitations

Data may not be in sync between explorations and reporting.

Offline events

When a user's device goes offline (for example, a user loses their internet connection while browsing your mobile app), Google Analytics stores event data on their device and then sends the data once their device is back online. Google Analytics ignores events that arrive more than 72 hours after the events are triggered.

Features with non-standard data processing

The following Google Analytics 4 features are not covered by the Analytics 360 data-processing service level agreement:

Feature Estimated processing time
Realtime Less than 1 minute
Unsampled exploration 1 hour
Attribution 4-8 hours
User-lifetime technique 24 hours
Insights 24 hours
Expanded data sets 24 hours
BigQuery daily export 24 hours after midnight in the property timezone
Data Import 24-48 hours

Features not covered by Reporting SLA

The following features are not covered by the Analytics 360 Reporting service level agreement:

Feature Availability
Realtime Best effort
Data API queries Best effort
Audience List API queries Best effort
PDF export Best effort
Unsampled exploration Best effort
Intelligence services (Insights, Single Search) Best effort
Queries that join data from other Google products (such as Google Ads) Best effort
Queries that join data that has been imported Best effort
Reporting queries with more than 1M rows of data Best effort
Exploration queries with more than 1M events Best effort

Report queries that have filters applied

(includes any filtered query from the reporting or exploration interface or from the Data API)

Best effort

Was this helpful?

How can we improve it?
true
Search
Clear search
Close search
Google apps
Main menu