Google Analytics updates data analysis and reporting

Google Analytics yesterday introduced a series of significant updates aimed at enhancing data analysis and reporting capabilities for users. These new features, announced just one day ago, provide tools for improved visualization, anomaly detection, and transaction tracking.

According to the official Google Analytics documentation, one of the key additions is the ability to visualize up to five rows of data directly within detail reports using a feature called plot rows. This functionality allows users to select specific rows and generate lines for each, facilitating quick comparisons and analysis.

Anomaly detection has been expanded to include detail reports, enabling users to uncover hidden patterns and potential issues in their data. The system now flags unusual data fluctuations, helping analysts identify unexpected traffic spikes, sudden drops, and other noteworthy trends. This feature utilizes a Bayesian state-space time series model to predict metric values and identify anomalies.

For hourly anomaly detection, the system uses a two-week training period. Daily anomaly detection employs a 90-day training period, while weekly anomaly detection uses a 32-week training period. Additionally, Google Analytics now offers anomaly detection across multiple metrics and dimension values at a single point in time, using principal components analysis (PCA) and cross-validation techniques.

The Events report has been enhanced with the ability to mark key events. Users can now identify their most important events by using the three-dot menu at the end of event rows. This feature allows for easier tracking and analysis of critical user interactions on websites or applications.

A new Transaction ID report has been introduced, providing granular insights into revenue streams. This report displays detailed information about each e-commerce transaction, including purchase completion rates, number of items purchased, and total revenue generated. The Transaction ID report can be accessed in the Monetization topic of the Life cycle collection or the Drive online sales topic of the Business objectives collection.

To populate the Transaction ID report, users need to send the event-level 'transaction_id' parameter with online sales events. The report includes dimensions such as Transaction ID and metrics like Ecommerce purchases, Ecommerce purchase quantity, and Purchase revenue.

Google has also made changes to the scope of acquisition reports. The User acquisition report now focuses on user-related metrics such as Total users, New users, and Returning users. Conversely, the Traffic acquisition report emphasizes session-related metrics, including Sessions, Engaged sessions, and Sessions per event.

Additional improvements include

1. Report descriptions: Users can now hover over any report title to reveal a brief description and a helpful link to learn more about the report.
2. Scope changes in acquisition reports: The User acquisition report now focuses on user-related metrics, while the Traffic acquisition report emphasizes session-related metrics.
3. New Transaction ID report: This report provides detailed information about each transaction using the transaction_id parameter.

These updates to Google Analytics reflect the platform's ongoing efforts to provide more comprehensive and accessible data analysis tools. By introducing features like plot rows and expanded anomaly detection, Google aims to help users identify trends and make data-driven decisions more effectively.

The new Transaction ID report, in particular, offers e-commerce businesses a more detailed view of their sales performance. By tracking individual transactions, companies can gain insights into purchasing patterns, revenue generation, and overall sales effectiveness.

Key facts

• Announced on September 3, 2024
• Plot rows feature allows visualization of up to 5 rows in detail reports
• Anomaly detection expanded to detail reports
• New Transaction ID report introduced for e-commerce insights
• Events report now allows marking of key events
• Acquisition reports have been updated to focus on user-related and session-related metrics separately
• Report descriptions added for easier understanding
• Anomaly detection uses different training periods for hourly (2 weeks), daily (90 days), and weekly (32 weeks) data
• PCA and cross-validation techniques used for multi-dimensional anomaly detection