GA4 vs Universal Analytics: Everything That Changed

The Core Difference: Events vs Sessions

Universal Analytics used a session-based model. Every interaction (pageview, event, transaction) lived within a session. Reports were organized around sessions and pageviews.

GA4 uses an event-based model. EVERYTHING is an event — including pageviews (now called page_view). There are no more "hit types." This makes GA4 more flexible for apps and complex user journeys but changes how you interpret basic reports.

What Got Renamed

| UA | GA4 |

|-----|-----|

| Bounce Rate | Engagement Rate (inverted) |

| Goals | Conversions |

| Custom Dimensions (Hit-level) | Event Parameters |

| Custom Dimensions (Session-level) | Removed |

| Custom Dimensions (User-level) | User Properties |

| View | Data Stream (per platform) |

What Got Removed

• View filters: UA allowed you to create filtered views (e.g., exclude internal traffic). GA4 has no equivalent "views" — use data filters instead.
• Average session duration: Replaced by engagement time and engaged session calculations.
• Bounce rate in default reports: It is still available but not prominently shown.

What Is New in GA4

User ID and Google Signals: GA4 can stitch together cross-device journeys if users log in or have Google account activity.

BigQuery export: Now free for all GA4 properties. Export raw event data for custom SQL analysis.

Predictive audiences: Machine learning-powered audiences that predict purchase probability, churn, etc.

Path Exploration and Funnel Exploration: New report types not available in UA.

The Reporting Lag

GA4 processes data more slowly than UA. Expect a 24-48 hour lag for standard reports. BigQuery exports can be near real-time but require setup.

Building Equivalent UA Reports in GA4

• Source/Medium report: Acquisition → Traffic acquisition, then change primary dimension to "Session source / medium"
• Landing pages: Engagement → Landing page
• Conversion by source: Advertising → Attribution report, or use Exploration