ANALYTICS GLOSSARY 70+ DEFINITIONS A-Z

Active Users

Active Users is the primary user metric in GA4. An active user is someone who has had an engaged session or when Analytics collects the first_visit event or engagement_time_msec parameter. In Universal Analytics, "Users" (total number of users) was the default; in GA4, "Active Users" is the default when you see "Users" in reports. This gives a more realistic picture of your actual audience, since passive visitors who immediately bounce count less heavily.

Attribution

Attribution is the process of assigning value to the various touchpoints a customer goes through before converting. For example: someone first clicks on a Google Ads ad, returns a week later via organic search, and ultimately converts through an email. Which channel gets the credit? That depends on your attribution model. GA4 uses data-driven attribution by default, where machine learning distributes value based on each touchpoint's actual contribution.

Attribution Model

An attribution model is the set of rules that determines how conversion value is distributed across touchpoints. The most well-known models are: last-click (all credit to the last channel), first-click (all credit to the first channel), linear (equal distribution), time decay (more credit to more recent touchpoints), and data-driven (distribution based on machine learning). GA4 still supports data-driven and paid-and-organic last-click. The chosen model has a significant impact on how you evaluate channels and allocate your budget.

Attribution Window

The attribution window (also called lookback window) determines how far back in time a touchpoint can still receive credit for a conversion. In GA4, you can set this to 30, 60, or 90 days for acquisition conversions. If your window is set to 30 days and someone first clicked on your ad 45 days ago, that initial touchpoint no longer counts in the attribution. For B2B with longer purchase cycles, a longer window is typically more advisable.

Average Engagement Time

Average Engagement Time is the GA4 successor to "Average Session Duration" from Universal Analytics. It measures the average time your website or app was actually in the foreground and actively being viewed. This is more accurate than the old session average, because it stops counting as soon as a user switches to another tab or locks their screen. A page with 2 minutes of average engagement time typically performs better than a page with 15 seconds.

Average Session Duration

Average Session Duration is the average length of all sessions on your website. In Universal Analytics, this was calculated as total session duration divided by the number of sessions — but the duration of the last pageview was not measured, leading to underestimation. In GA4, this metric has been replaced by Average Engagement Time, which measures more accurately. If you still encounter this metric in older reports: a session duration under 30 seconds often indicates irrelevant traffic or technical issues.

BigQuery

BigQuery is Google's cloud data warehouse that you can connect to GA4 for unlimited storage of raw event data. The free GA4 integration exports all your events to BigQuery daily, where you can query them with SQL. This is essential for advanced analyses that the standard GA4 interface doesn't support — think user-level path analyses, custom attribution models, or combining analytics data with CRM data. The free tier of BigQuery is more than sufficient for most small to mid-sized business websites.

Bounce Rate

The bounce rate is the percentage of sessions where a visitor leaves your website without any interaction. In Universal Analytics, a bounce was a session with only one pageview. In GA4, the definition has changed: a bounce is now a session that doesn't qualify as "engaged" — meaning shorter than 10 seconds, without a conversion, and with at most one screen view. Average bounce rates vary by industry: blogs sit around 70-80%, e-commerce around 40-50%, and landing pages around 60-70%. A high bounce rate isn't necessarily bad — on a contact page with just a phone number, it's expected. Learn more about optimizing your organic visibility to attract quality traffic.

Channel

A channel is a grouped category of traffic sources in Google Analytics. The default channel groups are: Organic Search, Paid Search, Direct, Referral, Social, Email, Display, and Affiliates. GA4 adds Organic Social, Paid Social, Organic Video, and Paid Video. Channels help you understand at a high level where your visitors are coming from. In GA4, you can also create custom channel groups to use your own categorization — useful when you want to bundle specific campaign types.

Client ID

The Client ID is a unique identifier that Google Analytics assigns to a browser cookie to recognize an individual browser/device combination. It is stored in the _ga cookie and looks like: GA1.2.123456789.1234567890. Important: a Client ID identifies a browser, not a person. If the same person visits your website on their laptop and phone, those are two different Client IDs. That's why the number of "users" in analytics is always an overestimate of the actual number of people.

Cohort Analysis

A cohort is a group of users who share a common characteristic within a certain period. In cohort analysis, you group users based on their first visit date and track their behavior in the following weeks. For example: "Of the users who first visited in week 1, what percentage returned in weeks 2, 3, and 4?" This reveals retention patterns that you don't see in standard reports. GA4 offers a built-in cohort exploration in the Explore section.

Consent Mode

Consent Mode is Google's system for adjusting analytics and advertising tags based on users' cookie consent. When a visitor declines cookies, Consent Mode sends anonymized pings to Google instead of full data. Google then uses modeling to estimate the missing data. There are two variants: Basic (tags are completely blocked without consent) and Advanced (cookieless pings are sent). Since March 2024, Consent Mode v2 is required for advertisers in the EEA.

Conversion

A conversion is a valuable action you define that a user performs on your website. In GA4, you mark any desired event as a conversion — for example, a form submission, purchase, or phone call. The difference from Universal Analytics: there it was called "Goals" and you were limited to 20. In GA4, you can set up to 30 conversion events per property. The conversion rate (the percentage of sessions that lead to a conversion) is one of the most important KPIs for any online marketing strategy.

Conversion Path

A conversion path is the complete series of touchpoints a user goes through before converting. In GA4, you can find this under Advertising > Conversion Paths. The report shows which channels, sources, and campaigns played a role in the path to conversion. Typical example: Organic Search (first contact) > Direct (return visit) > Email (conversion). By analyzing conversion paths, you discover which channels often serve as "assistants" but rarely get the direct conversion credit — crucial for fair budget allocation.

Cookie

A cookie is a small text file that your website stores in a visitor's browser. In analytics, first-party cookies are used to recognize users upon return. Google Analytics places the _ga cookie (valid for 2 years) and the _ga_<container-id> cookie. Due to privacy legislation (GDPR) and browser restrictions (Safari's ITP limits cookies to 7 days), cookie-based tracking is becoming increasingly unreliable. That's why Google is investing in cookieless measurement solutions via Consent Mode and machine learning modeling.

Cross-Domain Tracking

Cross-domain tracking ensures that a user navigating from one domain to another is recognized as the same user. Without cross-domain tracking, a visitor going from shop.yourcompany.com to www.yourcompany.com appears as a new user — and the session is incorrectly registered as "Referral." In GA4, you configure this via Admin > Data Streams > Configure Tag Settings > Configure Your Domains. It works by passing the Client ID as a parameter in the URL during domain transitions.

Custom Dimension

A custom dimension is a dimension you define that doesn't exist by default in Google Analytics. It allows you to add extra context to your data. Examples: logged in/not logged in, customer segment (prospect vs. customer), content category, or A/B test variant. In GA4, you set up custom dimensions at event-scope (linked to specific events) or user-scope (linked to the user). You can create up to 50 event-scoped and 25 user-scoped custom dimensions per property.

Custom Metric

A custom metric is a self-defined numeric value you can send along with events. Where custom dimensions describe categories (text), custom metrics describe quantities (numbers). Think of: scroll depth in percentages, number of products in shopping cart, or reading time in seconds. In GA4, you can create up to 50 custom metrics. They are always linked to an event parameter that you send via your Data Layer or gtag configuration.

Dashboard

A dashboard is a visual overview of your most important metrics and KPIs on a single screen. In GA4, the Home section provides an automatic dashboard, but most professionals build their own dashboards in Looker Studio (formerly Google Data Studio). A good dashboard shows at a glance: traffic, conversions, revenue, and trends compared to the previous period. The golden rule: if you need more than 30 seconds to interpret your dashboard, it's too complex.

Data Layer

The Data Layer is a JavaScript object (window.dataLayer) that makes structured information about your page and user interactions available to Google Tag Manager. It acts as an intermediary between your website and your tracking tags. Instead of hardcoding data directly into tags, you push information to the Data Layer — such as transaction data, product information, or form data. GTM reads this data and forwards it to Analytics, Google Ads, or other platforms. This makes your tracking maintainable, flexible, and less dependent on web developers.

Data-Driven Attribution

Data-driven attribution is an attribution model in GA4 that uses machine learning to distribute conversion value across touchpoints based on their actual contribution. Instead of fixed rules (like "everything to the last click"), the model analyzes which channel combinations actually lead to conversions compared to paths that don't. This is the default attribution model in GA4 and provides the fairest picture — provided you have sufficient conversion data (minimum 300 conversions and 3,000 click interactions per 30 days for optimal modeling).

Data Retention

Data retention in GA4 determines how long user and event-level data is stored for explorations and detailed reports. The options are 2 months (default) or 14 months. This only applies to data in the Explore section and freely available reports — aggregated data in standard reports remains available indefinitely. Tip: set data retention to 14 months immediately in Admin > Data Settings > Data Retention. And connect BigQuery for unlimited storage of your raw data.

Data Stream

A data stream is the connection between your website or app and your GA4 property. Each GA4 property can have multiple data streams: one for your website, one for your iOS app, and one for your Android app. The web data stream contains your Measurement ID (starts with G-). Cross-platform data is automatically merged in GA4 within the same property, giving you a holistic view of user behavior across web and app.

Debug View

Debug View is a real-time tool in GA4 that lets you monitor and validate incoming events. You activate it via the Google Analytics Debugger Chrome extension or by adding debug_mode: true to your gtag configuration. In Debug View, you see exactly which events are coming in per user, with which parameters, and at what moment. It's indispensable when implementing or troubleshooting new tracking — you can see immediately whether everything works correctly before your data goes live.

Dimension

A dimension is a descriptive attribute of your data — the "what" or "who." Examples: page title, city, device category, source/medium, or browser. Dimensions are categorical (text) as opposed to metrics (numbers). In every GA4 report, you combine dimensions with metrics: "how many sessions (metric) per city (dimension)" or "what is the conversion rate (metric) per channel (dimension)." Understanding the difference between dimensions and metrics is fundamental to any analytics analysis.

Direct Traffic

Direct traffic is traffic where Google Analytics cannot determine source information. This happens when someone types your URL directly into the browser bar, uses a bookmark, or clicks a link in an untagged email or PDF. In practice, "Direct" is often a catch-all for traffic with missing referrer information. That's why it's important to consistently tag all your campaigns with UTM parameters. An unusually high percentage of direct traffic (above 30-40%) often indicates poor campaign tagging rather than genuine direct visitors.

E-commerce Tracking

E-commerce tracking is the implementation of specific events in GA4 to measure the purchase process: product views (view_item), cart additions (add_to_cart), checkout steps (begin_checkout), and transactions (purchase). GA4 provides standard e-commerce reports showing revenue, products, average order value, and funnel drop-off. The implementation requires a well-structured Data Layer that sends product information (name, price, category, quantity) with each event.

Engaged Session

An engaged session is a session that meets at least one of these conditions: lasts longer than 10 seconds, contains a conversion event, or contains two or more screen views/pageviews. This is a new concept in GA4 that is directly related to the revised bounce rate. The engagement rate (percentage of engaged sessions) is the inverse of bounce rate: a 65% engagement rate means a 35% bounce rate. The 10-second threshold is configurable in Admin > Data Streams > Configure Tag Settings.

Event

An event is the basic unit of data collection in GA4. Everything is an event: pageviews (page_view), clicks, scrolls, video views, file downloads, and conversions. Each event can carry up to 25 parameters that provide additional context. GA4 automatically collects certain events (first_visit, session_start, page_view), while you can toggle enhanced measurement events (scroll, outbound click, file download) on or off. Additionally, you can create custom events for interactions specific to your business — such as opening a chat widget or viewing a pricing page.

Event Parameter

An event parameter is a piece of additional information sent along with an event. With the page_view event, GA4 automatically sends parameters like page_title, page_location, and page_referrer. With a custom event like form_submit, you can add your own parameters: form_name, form_type, form_location. To use parameters in your standard reports, you need to register them as a custom dimension or custom metric in the GA4 admin.

Exit Rate

The exit rate is the percentage of sessions that end on a specific page, calculated as the number of exits divided by the number of pageviews for that page. The difference from bounce rate: exit rate applies to all sessions (including sessions with multiple pageviews), while bounce rate only applies to single-page sessions. A high exit rate on your thank-you page or confirmation page is perfectly normal and even desirable. A high exit rate on a product page or checkout step is a signal that something is wrong.

Exploration

Exploration is the advanced analysis section of GA4 (formerly "Analysis Hub"). Here you build custom reports using techniques like: Free form (pivot tables), Funnel exploration (funnel analysis), Path exploration (path analysis), Segment overlap, User lifetime, and Cohort exploration. Explorations give you much more flexibility than standard reports, but are limited to the configured data retention period (2 or 14 months). For long-term analyses, use BigQuery.

Filter

A filter restricts or modifies the data that appears in your Analytics reports. In GA4, filters are more limited than in Universal Analytics: you can set up data filters to exclude internal traffic (based on IP address) or developer traffic. For more advanced filtering, use comparisons in reports or segments in explorations. In Universal Analytics, you could make disastrous mistakes with filters (permanently deleting data); in GA4, that risk is smaller because filters work non-destructively — the raw data remains intact.

First-Party Data

First-party data is information you collect yourself through your own channels: website analytics, CRM data, email lists, purchase history, and customer feedback. In contrast to third-party data (purchased from external parties) and second-party data (shared by partners), first-party data is the most reliable and GDPR-compliant. With the disappearance of third-party cookies, first-party data is becoming increasingly valuable for targeting and personalization.

Funnel

A funnel is a visual representation of the steps users go through toward a conversion, showing the drop-off percentage at each step. In GA4, you build funnels in Explore > Funnel Exploration. You define the steps (for example: homepage > product page > shopping cart > checkout > purchase) and immediately see where the most visitors drop off. An open funnel counts users who enter at any step; a closed funnel only counts users who start at step 1. Funnel analysis is one of the most powerful tools for prioritizing conversion rate optimization.

GA4 (Google Analytics 4)

GA4 is the current version of Google Analytics, launched in October 2020 and the only version since July 2023 after Universal Analytics was discontinued. The biggest changes: a fully event-based data model (sessions are no longer the basic unit), cross-platform tracking (web + app in one property), built-in machine learning predictions, privacy-first design with Consent Mode, and free BigQuery integration. The interface has changed significantly and many marketers have a learning curve — but the analytical capabilities are considerably more powerful.

Goal

A goal was the concept in Universal Analytics for defining conversions — with a maximum of 20 per view. In GA4, goals no longer exist; they have been replaced by conversion events. You simply mark any event as a conversion with a toggle. This is more flexible: you can set up to 30 conversion events, and you're not bound to the four goal types (destination, duration, pages/screens, event) of Universal Analytics.

Google Signals

Google Signals is a GA4 feature that uses data from users logged into their Google account for cross-device reporting and remarketing. When activated, GA4 can recognize that the same person visits your website on multiple devices. This improves the accuracy of your user counts and demographic reports. Note: with low traffic, activating Google Signals can lead to data thresholding — GA4 then hides rows to protect individual user privacy.

Google Tag Manager (GTM)

Google Tag Manager is a free tag management system from Google that lets you manage tracking codes (tags) on your website without modifying the source code. You place the GTM container on your website once, and then manage all tags — Google Analytics, Google Ads conversion tracking, Facebook Pixel, Hotjar, and more — through the GTM interface. GTM works with three concepts: tags (code that executes), triggers (when the tag fires), and variables (dynamic values). It makes marketers more independent from developers and is the standard for professional tracking implementations.

gtag.js (Global Site Tag)

gtag.js is Google's JavaScript library for directly implementing Google Analytics, Google Ads, and other Google products on your website without Tag Manager. You place the gtag code in the <head> of your page and configure everything via JavaScript calls. While gtag.js works fine for simple implementations, most professionals choose Google Tag Manager for its flexibility, version control, and the ability to avoid modifying website code for every change.

Hit

A hit was the generic name in Universal Analytics for every interaction that sent data to the Analytics server: pageview hit, event hit, transaction hit, social hit. In GA4, the hit concept has been replaced by the unified event model — everything is now an event. The term is still commonly used in the industry, but technically, "event" is the correct GA4 terminology. In UA, the free version had a limit of 10 million hits per month per property.

Internal Traffic

Internal traffic is traffic from your own employees, developers, or agency partners that you want to exclude from your analytics data to prevent skewing. In GA4, you define internal traffic rules via Admin > Data Streams > Configure Tag Settings > Define Internal Traffic, where you specify IP addresses or ranges. Then you activate a data filter to exclude this traffic. Always test with the filter set to "Testing" first before switching to "Active" — otherwise you risk permanent data loss.

Key Event

Key Event is the new name for what was previously called "Conversion" in GA4 (renamed in March 2024). Google made this name change to avoid confusion with the term "Conversions" in Google Ads, which has a different definition. Functionally, nothing has changed: you mark events as key events and they appear in your reports as important actions. In the Google Ads integration, imported GA4 key events are automatically called "Conversions." Confusing? Yes. But the data is identical.

KPI (Key Performance Indicator)

A KPI is a measurable value that indicates how effectively you're achieving a business objective. In analytics, typical KPIs include: conversion rate, revenue per session, cost per acquisition, bounce rate, and average order value. The difference from a metric: every KPI is a metric, but not every metric is a KPI. Pageviews are a metric; pageviews per session on your product pages becomes a KPI only when it's directly tied to a business goal. Choose a maximum of 5-7 KPIs for your dashboard — more leads to analysis paralysis.

Landing Page

The landing page is the first page a user visits when starting a session. In GA4, you find landing page data under Reports > Engagement > Landing Page. This report is crucial: it shows which pages bring in visitors, and how those pages perform in terms of bounce rate, conversion rate, and engagement. A product page with lots of organic landings but a high bounce rate is a direct optimization opportunity. The landing page dimension in GA4 shows the path (/products/shoes) without the domain name.

Last-Click Attribution

Last-click attribution is an attribution model that assigns 100% of the conversion value to the last channel the user interacted with before converting. It's the simplest model and was the standard for years. The downside: it ignores all preceding touchpoints that helped the customer on the path to conversion. GA4 still offers "Paid and organic last click" as an alternative to data-driven attribution. Last-click typically overvalues branded search and direct traffic, and undervalues upper-funnel channels like display and social.

Looker Studio (formerly Google Data Studio)

Looker Studio is Google's free business intelligence and dashboarding tool. You connect data sources — GA4, Google Ads, Google Search Console, Google Sheets, BigQuery, and more — and build interactive visual reports. It's the standard tool for analytics dashboards at agencies and in-house teams. Advantages over the GA4 interface: you combine data from multiple sources, have full control over visualization and layout, and can share reports with stakeholders who don't have GA4 access. At Searchlab, we use Looker Studio for all client reporting.

Measurement ID

The Measurement ID is the unique identifier for your GA4 web data stream, in the format G-XXXXXXXXXX. You use this ID to connect your website to your GA4 property — either via gtag.js or via Google Tag Manager. It replaces the "UA-" tracking ID from Universal Analytics. You can find it in GA4 under Admin > Data Streams > click on your web stream. A single GA4 property can have multiple Measurement IDs if you track multiple websites or subdomains.

Measurement Protocol

The Measurement Protocol is an API that lets you send events directly to GA4 via HTTP requests, without a browser or JavaScript. This is useful for server-side tracking: measuring offline conversions, kiosk interactions, IoT devices, or sending CRM data back to Analytics. You send events with an API secret and your Measurement ID. Important difference from client-side tracking: Measurement Protocol hits are not blocked by ad blockers or cookie consent, but you are responsible for respecting privacy preferences.

Metric

A metric is a quantitative measurement of your data — the "how many." Examples: sessions, users, pageviews, bounce rate, conversion rate, revenue. Metrics are always numbers (whole, decimal, or percentage) as opposed to dimensions (text/categories). In every report, you combine metrics with dimensions: "how many sessions (metric) per country (dimension)." GA4 has dozens of built-in metrics, and you can add your own custom metrics.

New vs. Returning Users

New vs. returning users is a segmentation that distinguishes between users visiting your website for the first time and users who are returning. In GA4, this is determined based on the first_visit event and the Client ID in the cookie. A typical split is 60-70% new and 30-40% returning, but this varies significantly by website type. Keep in mind that this metric is becoming increasingly unreliable: if someone clears their cookies, uses a different browser, or surfs in incognito mode, they are incorrectly counted as "new."

Not Provided

(not provided) is the value Google Analytics shows for organic search terms that Google does not share due to search result encryption (SSL). Since 2011, Google has been gradually hiding search terms from logged-in users; by now, 95%+ of all organic search terms are "(not provided)." To still gain insight into your organic search terms, use Google Search Console (free) and connect it to GA4.

Organic Search

Organic Search is traffic that comes through unpaid search results on search engines like Google, Bing, and DuckDuckGo. In GA4, it's one of the default channel groups. Organic search is recognized because the referrer is a known search engine and no paid advertising campaign is linked to it. It's often the most important channel for sustainable growth. Investing in SEO structurally increases your organic search traffic without paying per click as with Google Ads.

Organic Social

Organic Social is traffic from social media platforms (LinkedIn, Facebook, Instagram, X) that does not come from paid ads. GA4 splits social traffic into Organic Social and Paid Social, which is an improvement over the combined "Social" channel in Universal Analytics. This way, you can directly see the difference between your organic social media efforts and your advertising investments. Correctly tagging paid campaigns with UTM parameters is essential to keep this split accurate.

Pageview

A pageview is triggered every time a page loads in the browser. In GA4, a pageview is the page_view event, which is automatically collected. Difference from unique pageviews (UA): GA4 no longer has a "unique pageviews" metric. Instead, you use "Views" (which combines pageviews and screen views) or analyze at the user level in explorations. Pageviews are the most basic metric but say little about quality — 10,000 pageviews with a 90% bounce rate are worth less than 3,000 pageviews with a 60% engagement rate.

Path Exploration

Path exploration is an analysis technique in GA4 Explore that lets you visualize which pages or events users visit sequentially. You can look forward (which pages do users visit after the homepage?) or backward (which path led to the contact page?). This reveals unexpected navigation patterns, dead ends, and popular routes that you don't see in standard reports. Path exploration is particularly valuable for optimizing your website architecture and internal linking.

Predictive Metrics

Predictive metrics are machine learning-generated predictions in GA4. Three are available: Purchase probability (likelihood an active user will purchase in the next 7 days), Churn probability (likelihood an active user won't return in the next 7 days), and Predicted revenue (expected revenue from a user in 28 days). To activate these metrics, you need at least 1,000 returning users per week who do/don't convert. They're powerful for building audiences in Google Ads — for example: target users with high purchase probability with a special offer.

Property

A property is the container level in Google Analytics where your data is collected and reported. In GA4, a property is the level directly under your Account. Each property contains one or more data streams (web, iOS, Android). The difference from Universal Analytics: in UA, you had Account > Property > View. In GA4, Views no longer exist — you only have Account > Property. A single property can combine data from your website, iOS app, and Android app, enabling cross-platform analysis.

Realtime Report

The realtime report in GA4 shows activity on your website in the past 30 minutes. You can see the number of active users, which pages they're viewing, where they came from, which events they're triggering, and how many conversions are taking place. It's useful for monitoring campaign launches, verifying new tracking implementations, and following promotions in real time. Note: realtime data may slightly differ from final report data due to processing delays.

Referral

Referral is traffic that comes through a click on a link on another website. In GA4, referral traffic is automatically recognized based on the HTTP referrer header. You see the specific referring domain as the source and "referral" as the medium. Unwanted referral traffic — such as payment gateways, your own subdomains, or spam — can be excluded in Admin > Data Streams > Configure Tag Settings > List Unwanted Referrals. This prevents sessions from being incorrectly split or misattributed.

Regex (Regular Expression)

Regex (regular expression) is a pattern description for matching text. In analytics, you use regex for filtering, segmenting, and matching URLs, page titles, and event names. Examples: /blog/.* matches all blog URLs, ^/contact$ matches exactly the contact page, and (new-york|los-angeles|chicago) matches URLs containing one of these cities. GA4 uses regex in explorations and audience definitions. In Looker Studio and Google Tag Manager, regex is equally indispensable. The investment in learning basic regex pays for itself dozens of times over in analytics efficiency.

Remarketing Audience

A remarketing audience is a target group you build in GA4 based on website behavior and then share with Google Ads or other advertising platforms. Examples: visitors who viewed a product but didn't purchase, users who visited at least 3 pages, or people who watched a video for more than 50%. GA4 offers a powerful audience builder with conditions, sequences, and exclusions. The audience is automatically populated and updated, ensuring your ads always reach the most relevant group.

Sampling

Sampling is analyzing a sample of your data instead of the full dataset to load reports faster. GA4 applies sampling when an exploration contains more than 10 million events — you'll then see a green, yellow, or red icon indicating quality. Standard reports in GA4 are not sampled (they use pre-aggregated data). To avoid sampling in advanced analyses: use BigQuery for unsampled queries, limit your date range, or make your segments/filters more specific.

Segment

A segment is a subset of your data based on specific criteria. In GA4, you use segments in Explorations to isolate and compare groups of users, sessions, or events. Three types: user segment (all sessions from users who meet a condition), session segment (only the sessions that meet the condition), and event segment (only the events that match). For example, compare "users who purchased" with "users who didn't purchase" to discover behavioral differences that lead to conversion optimization.

Server-Side Tagging

Server-side tagging is an implementation method where tracking tags run on a server instead of in the visitor's browser. Instead of the browser sending data directly to Google Analytics, Google Ads, and Facebook, the browser sends data to your server (via a server-side GTM container), which then forwards it to the appropriate platforms. Benefits: better data quality (less loss from ad blockers), faster website (fewer scripts in the browser), more control over what data you share, and longer cookie lifespan (first-party server-set cookies). The downside: it costs money (cloud hosting) and is more complex to set up.

Session

A session is a group of user interactions on your website within a specific period. In GA4, a session starts with the session_start event and ends after 30 minutes of inactivity (configurable up to 7 hours and 55 minutes). Unlike Universal Analytics, a GA4 session is not restarted at midnight or when the campaign source changes. This means GA4 typically reports fewer sessions than UA for the same website. The session remains the cornerstone of many analytics analyses, despite GA4's event-first approach.

Session Source/Medium

Session source/medium indicates where a session came from. The source is the specific origin (google, facebook, newsletter) and the medium is the channel type (organic, cpc, email, referral). Together they form the dimension "Session source / medium" — for example: google / organic, facebook / cpc, newsletter / email. In GA4, there are two variants: session-scoped (source of the session) and user-scoped (first source through which a user ever arrived). Pay attention to this difference: it fundamentally affects your analyses.

Scroll Tracking

Scroll tracking measures how far users scroll down on a page. GA4 offers automatic scroll tracking via enhanced measurement: the scroll event triggers when a user reaches 90% of the page. For more detailed insight (25%, 50%, 75%, 90%), implement custom scroll tracking via Google Tag Manager with a scroll depth trigger. Scroll data is valuable for content optimization — if most readers drop off at 30% of a blog post, the introduction probably isn't engaging enough or the content is too long for the topic.

Tag

A tag is a piece of code (JavaScript) that collects data and sends it to an analytics platform. In Google Tag Manager, you manage tags centrally: Google Analytics tags, Google Ads conversion tags, Facebook Pixel, LinkedIn Insight Tag, Hotjar, and more. Each tag has a trigger (when should the tag fire) and optionally variables (what data gets sent). The advantage of tags via GTM: you can add, modify, and remove them without changing your website's source code — and with built-in version control, you can always roll back if something goes wrong.

Thresholding

Thresholding is GA4's privacy measure where rows are hidden from reports when the numbers are too low to protect individual users. You can recognize it by a triangular warning icon in your report. It occurs more frequently when Google Signals is activated, with small data populations, or in combination with demographic data. Solutions: expand your date range, remove granular dimensions, deactivate Google Signals (if you're not using it for cross-device), or export your data via BigQuery where thresholding doesn't apply.

Tracking Code

The tracking code is the piece of JavaScript you place on your website to collect data for Google Analytics. In GA4, this is either the gtag.js snippet (with your Measurement ID) or the Google Tag Manager container (with your GTM ID). The code must be on every page, preferably in the <head> section, to ensure visits are measured before the visitor clicks away. Without a correctly implemented tracking code, you collect no data — it's literally the foundation of your entire analytics setup.

Trigger

A trigger in Google Tag Manager defines when a tag should execute. There are dozens of trigger types: pageview, DOM ready, click on an element, form submission, scroll depth, YouTube video interaction, custom event, and more. You combine triggers with filters to make them fire precisely: "Fire on all clicks on links starting with mailto:" or "Fire on page_view when the URL contains /thank-you." Incorrect triggers are the number one cause of tracking problems — always test via Debug View or GTM Preview Mode.

Unique Users

Unique users is the estimated number of individual people who visited your website. In GA4, the primary metric is called "Active Users" when you see "Users" in reports. The count is based on cookies (Client ID) or Google Signals (if activated). Due to cookie restrictions, incognito mode, and multiple devices, the actual number of unique people is always lower than what Analytics reports. For a business website with 10,000 "unique users" per month, the real number might be 7,000-8,000 people. Use User-ID tracking to count authenticated users more accurately.

User ID

User ID is a feature in GA4 that lets you assign your own persistent identifier to logged-in users. This allows GA4 to recognize the same person across multiple browsers, devices, and sessions — which isn't possible with cookies alone. You implement User ID by pushing your own ID (never personally identifiable information!) to the Data Layer at login. GA4 then uses the identity space hierarchy: User ID > Google Signals > Device ID to compile the most accurate user reporting.

UTM Parameters

UTM parameters are tags you add to URLs to track the source of campaign traffic in Google Analytics. There are five UTM parameters:

• utm_source — the specific source (google, linkedin, newsletter)
• utm_medium — the channel type (cpc, email, social, referral)
• utm_campaign — the campaign name (spring-sale-2026, product-launch)
• utm_term — the keyword (optional, for paid search ads)
• utm_content — the variant (optional, for A/B tests of the same campaign)

Example URL: https://yoursite.com/?utm_source=linkedin&utm_medium=social&utm_campaign=analytics-blog. Use consistent naming conventions (always lowercase, hyphens for spaces) and document your conventions in a spreadsheet. Without UTM parameters, your campaign traffic ends up in "Direct" — a black box you can't analyze. Learn more about effective campaign tracking in our AI marketing agency approach.

Variable

A variable in Google Tag Manager is a dynamic value you can use in tags and triggers. There are two types: built-in variables (Page URL, Click URL, Form ID, Referrer, etc.) and user-defined variables (Data Layer variables, JavaScript variables, lookup tables, etc.). Variables make your tracking setup dynamic: instead of hardcoding a page URL in a tag, you use the {{Page URL}} variable. This makes your implementation reusable, maintainable, and less error-prone.

View

A view was a filtered perspective on your data within a property in Universal Analytics. Many professionals created three views: a raw view (unfiltered, as backup), a test view (to test filters), and a master view (production with all filters active). In GA4, views no longer exist. Instead, you use data filters (for internal traffic), comparisons (for ad-hoc filtering in reports), and segments (in explorations). Some professionals miss views, but the GA4 approach is less error-prone — in UA, you could accidentally permanently lose data through a misconfigured view filter.

Web Vitals (Core Web Vitals)

Web Vitals are performance metrics from Google that measure the user experience of your website. The three Core Web Vitals are: LCP (Largest Contentful Paint — load time of the largest element, target: <2.5 seconds), INP (Interaction to Next Paint — responsiveness, target: <200ms), and CLS (Cumulative Layout Shift — visual stability, target: <0.1). They are an SEO ranking factor and you can measure them via Google Search Console, PageSpeed Insights, and as custom events in GA4 (via the web-vitals JavaScript library). Good Web Vitals improve both your user experience and your organic visibility.