Google Analytics 4
Tracking a Website
- add a small piece of Javascript tracking code to each page on your site - this code will collect anonymous info about how a user interacts with a page
What else can the tracking code collect?
- info from the browser (i.e. language, type of browser, device and operating system) - traffic source (what brought the user to the site, such as search engines, ads, or emails)
Property
- the container for your reports based on data collected from apps and websites -multiple can be applied to each account -each collects data independently of each other using unique tracking IDs that appear in tracking code
Google Signals
- uses data from users that are signed in to a google account - event data is associated with google accounts
Device ID
- websites: comes from users browser - apps: comes from app-instance ID - automatically used as an identity space
Views
-multiple can be applied to each property - narrow down where you want to see data from/focus on that data
Purchase Funnel
Acquisition - build awareness and acquire user interest Behavior - users engage with your business Conversion - user becomes customer and transacts with your business
Defining audiences
GA4 allows you to define audiences using data you collect from users (i.e. IDs that are sent to Analytics)
What is BigQuery?
a server-less, cost effective, multicloud data warehouse - designed to help analyze data through machine learning, geospatial analysis, and business intelligence
Conversion
a type of event that has been assigned a certain value (i.e. purchases or downloads) - often the most important events are marked as conversions - can be used in Google ads - contribute to the success of your business
Goals
a way to track conversions (aka business objectives) from your website - ex: how many users purchased a product or signed up for a newsletter
Google Tag Manager
add and update website tags without needing web development experience
Event Parameters
additional pieces of information sent with events that can further specify an action taken by a user, or add more context to an event
Analytics account structure
analytics account --> property --> data stream (an account contains 1 or more properties, and each property contains 1 or more data streams)
Unified website and app reporting
analytics allows you to measure apps and websites together in a common language across all reports - events and conversions are reported the same whether they happen in app or on the web
Full reporting across devices
analytics can measure a single user journey across devices based on user-identifiers
Attribution
assigning credit for conversions to different ads, clicks, and factors which helped the user complete a conversion
Ads-preferred model: Last click
attributes 100% conversion credit to the last Google ads a customer interacted with before converting - if there are no ads, cross channel model last click is used
User Properties
attributes about who is using your app/website that can help you understand your user base better
Cross channel model: Position-based
attributes conversion credit 40% to first and last interactions, and 20% evenly to middle interactions
Cross channel model: Data-driven attribution
attributes conversion credit based on data for each conversion type - your account data is used to calculate contribution of each interaction - each interaction receives some sort of credit
Cross channel model: Linear
attributes conversion credit equally across all channels the customer engaged with before converting
Cross channel model: First click
attributes conversion credit to the first channel the customer engaged with before converting
Cross channel model: Last click
attributes conversion credit to the last channel the customer engaged with before converting
Cross channel model: Time decay
attributes more conversion credit to interactions that happened closer in time to conversion - credit distributed using 7 day half life (i.e. an interaction 8 days before conversion gets half the credit an interaction 1 day before conversion would get)
What does a tracking code do when a page loads?
collects and sends updated info about the user's activity
Segment Overlap
compares up to 3 user segments - helps isolate specific audiences based on certain conditions
Roll-up property
contains data from 2 or more source properties - can include source data from ordinary properties and subproperties, but not other roll-up properties - used to combine products, brands, or regions (combines many source properties into a single roll-up property)
How can data be exported to BigQuery?
daily: full export of data takes place once a day streaming: continuous export of data throughout the day (some data isn't included right away due to processing times) - you can choose (geographic) location where you'd like data to be stored
What data model does GA4 use?
flexible event-based model - collects interactions as events (record things like pageviews, button clicks, user actions, etc) - processes each user interaction as a standalone event - good for measuring how people engage with your app or website
What type of reporting does GA4 use?
full cross-device and cross-platform reporting
Subproperty
gets its data from 1 other property called the source property - the data in subproperty is usually a subset of data from its source property - use for data and user governance (controls what data is included/what users can see the data)
Path exploration
helps you understand how users progress from one stage in customer journey to the next - unlike funnel explo., which follows a single defined path, path explo. can follow a number of undefined paths
Cohort exploration
lets you explore behavior of groups of users that share a common characteristic over time
What does BigQuery do?
lets you run SQL (sales qualified lead) queries - you can query large sets of data
What are tags?
little pieces of code that live on a webpage - when a user completes a certain action that you care about (i.e. adding a product to their shopping cart) the tag fires - aka pixels or snippets - helps you understand actions taken by users on a website
Data Stream
lives within a property and is a source of data from apps or websites
Global Site Tag
manage tags by coding javascript directly on your website
Metric
numeric value that represents your data
User ID
provides the system with a way of uniquely identifying a particular user amongst all the users of that system - data can be used to track user journeys across devices
Free form technique
represents data in a cross-tab layout
What does ROI mean
return on investment - a calculation that shows the value of return on investment related to that investment's costs
User Lifetime
shows how users have behaved since they first interacted with your site or app
Dimension
text-based label that represents your data - certain values are assigned
What happens when Analytics processes data?
the data is stored in a database where it cannot be changed - appears as reports
Conversion modeling
the use of machine learning to measure the impact of marketing efforts - used when a subset of conversions can't connect to ad interactions
Tracking in the Google Store
tracking code shows: - how many users visit a page (one product page vs another) - how many users bought an item by tracking whether they made it to the purchase confirmation page
Firebase
used to connect analytics to an app
Dynamic tagging
used to connect analytics with a website
Funnel exploration
visualizes the steps a user takes towards a key task or conversion - can see where users enter and leave funnels