Google Analytics 4 audit, part 1

Hello again from the most amazing mini degree from cxl for deep understanding of data analytics

Today is review number 3, with a start with new module, and this module is all about google analytics, it is history, new features and more

The first course in this module talking about google analytics 4 “GA4” audit 

GA4 is the newest version of google analytics, have a lot of new features with event base tracking not session tracking like before

Analytics structure :

Before digging into new features and implementation we need to understand first what the structure you need for your business

Small setups

For a business with one website, set up an Analytics account that contains a Google Analytics 4 property with

1 web data stream

For a business with one mobile-app game, set up an Analytics account that contains a Google Analytics 4 property with

1 app data stream for the Android game

1 app data stream for the iOS game

1 web data stream for the supporting website if there is one (e.g., marketing site)

Larger setups

To decide on the account structure for a larger business (e.g., multi-national, multi-brand), it's helpful to ask the following questions.

Account

Think of an account as a collection of properties whose data is owned by a single legal entity and is governed by region-specific terms of service.

Is it important that the data from each region is owned by a distinct legal entity within that region?

Yes. Create multiple accounts, one for each region.

No. Create one account in the region that houses your company headquarters.

Property

A property represents the data for one logical user base (e.g., users of a single cross-platform application), and is the most granular level at which Analytics processes data. A property is also the level of the Analytics account hierarchy at which you create integrations with other products (e.g., Google Ads).

Is the data you collect related to a single logical user base? When you link Analytics to other products, do you want to share that entire body of data with each product?

Yes. Create one property.

No. Create a separate property for each logical user base.

Data stream

A data stream is the source of data from an app or website. A data stream is also the level at which you can control data-collection features via the Firebase SDK or the global site tag.

App data stream: You can have one data stream for each combination of app-package name and platform.

Web data stream: In most cases, you should use a single web data stream to measure the web user journey. To ensure consistent user and session reporting for web journeys that span domains, use a single web data stream combined with cross domain measurement.

Data filters and roll-ups

Do you need to filter data so that you can restrict specific data to specific members of your organization? Do you need to roll up data from multiple properties into a single property?

Yes. Google Analytics 4 properties do not currently support reporting views or Roll-Up Properties.

No. You're all set.

Then what else, then we need to know difference between universal ga and ga4 for perfect implementation, and the main different is that GA4 is measuring hits type,  include page hits, event hits, ecommerce hits, and social interaction hits.

In contrast, Google Analytics 4 data is event-based, with the principle that any interaction can be captured as an event. As such, Universal Analytics property hit types translate to events in a Google Analytics 4 property.

Events

Events represent a fundamental data model difference between Universal Analytics and Google Analytics 4 properties.

A Universal Analytics event has a Category, Action, and Label and is its own hit type. In Google Analytics 4 properties, every "hit" is an event; there is no distinction between hit types. For example, when someone views one of your website pages, a page_view event is triggered. 

Google Analytics 4 events have no notion of Category, Action, and Label and, unlike Universal Analytics reports, Google Analytics 4 reports do not display Category, Action, and Label. Therefore, it’s better to rethink your data collection in terms of the Google Analytics 4 model rather than port your existing event structure to Google Analytics 4.

Example 

Universal analytics 

Example: Tracking Video Interactions

Video Play

Event Category = Video

Event Action = Play

Event Label = Dog Finds Favorite Bone

Event Value = Optional

Custom Dimension 1 = YouTube

Custom Dimension 2 = Charles Farina

Analytics 4

Example: Tracking Video Interactions

Video Play

Event Name = Video

Video Action = Play

Video Name = Dog Finds Favorite Bone

Video Type = YouTube

Video Author = Charles Farina

Session 

A session is a group of user interactions with your website that take place within a given time frame.

Google Analytics 4 session metrics are derived from the session_start event, an automatically collected event. The duration of a session is based on the time span between the first and last event in the session. This and other nuances can lead to sessionization differences between your Universal Analytics and Google Analytics 4 properties. 

Active user calculation

User activity is detected automatically in Google Analytics 4.  In contrast, Universal Analytics relies on manual instrumentation (firing of an interactive event). A user can launch an app and be considered an active user in Google Analytics 4 but not in Universal Analytics. This may lead to higher active user counts for Google Analytics 4.

Session counting

Some aspects of session counting in Google Analytics 4 differ from Universal Analytics. In Universal Analytics, a new campaign will start a new session regardless of activity; In Google Analytics 4, a new campaign does not begin a new session. This may lead to lower session counts in your Google Analytics 4 reports.

Late hits may also be a factor. In Universal Analytics, hits are processed if they arrive within 4 hours of the close of the preceding day. Google Analytics 4 processes events which arrive up to 72 hours late. This may lead to lower session counts in your Universal Analytics property -- especially for apps -- and may also lead to variation in reported figures within these 72 hours.

Logged Google Analytics 4 events are uploaded automatically when iOS apps are backgrounded. This is not the case in Universal Analytics. As a result, iOS-related metrics may be significantly higher in your Google Analytics 4 reports.

Custom dimensions/metrics

Custom dimensions and custom metrics in Universal Analytics are used to add information to collected data. In Google Analytics 4, event parameters serve this purpose. Map your custom dimensions and metrics as follows, according to their scope. Like Event scoped custom dimension, user scoped custom dimension and ecommerce parameters 

User id

User ID in Google Analytics 4 properties presents a cross-platform, cross-device view of how users interact with your app or website. To use this feature, you have to be able to generate your own unique, persistent IDs, consistently assign those IDs to your users, and include the IDs along with the data you send to Analytics. Analytics creates a single user journey from all the data that is associated with the same user ID. Unlike Universal Analytics, a Google Analytics 4 property incorporates User ID natively across all reporting, analysis and insights and does not require a separate User-ID reporting view.

Parameters

In Google Analytics 4 properties, you can send parameters with each event. Parameters are additional pieces of information that can further specify the action the user took, or add further context to the event. 

User property 

User properties are attributes about the users who interact with your app or website. They are used to describe segments of your user base, such as language preference or geographic location.

Event batching for Google Analytics 4 properties

In Google Analytics 4 properties, most events are batched client-side. However,

Conversion events are always transmitted immediately, although they may be part of a batch

Containers loaded in debug mode will never batch events, to facilitate the realtime DebugView

If any events are still held client-side when the user leaves the page (e.g. by navigating to another page), those events are sent immediately

In browser environments that do not support the sendBeacon API, all events are sent as they happen, with no batching.

In the next article we will dig more in google analytics reports and some advanced implementation