

Why Your GA4 Bounce Rate Is Just Engagement Rate Flipped
Your GA4 bounce rate is the share of sessions that were not engaged. How Google defines it, how it differs from Universal Analytics, and where to add it.
Browse definitions for analytics terms, privacy concepts, and measurement jargon in one crawlable index.
130 terms


Your GA4 bounce rate is the share of sessions that were not engaged. How Google defines it, how it differs from Universal Analytics, and where to add it.


Google's GA4 landing page report lists the first page of every session. Where to find it, its default metrics, why (not set) rows appear and how to add source.


Set up GA4 site search with enhanced measurement, add your own query parameters, find the Search term dimension and track searches without a URL.


Set GA4 user properties with gtag or Tag Manager, register them as user-scoped custom dimensions and stay inside the 25-property limit.


Learn how to link Search Console to GA4 in Admin, which roles you need, how to publish the two reports in Library and what the data can't join.


Chrome's third party cookie deprecation ran from a 2020 plan to a 2025 reversal. Every Google announcement, what Chrome does now, and what it means.


The bfcache, or back/forward cache, restores a page from memory on back and forward clicks. Which pages qualify, and why analytics miss or double count it.


Trackers used CNAME cloaking to pose as your subdomain. See how the DNS trick works, why Safari caps its cookies at 7 days and how Brave and uBlock uncloak it.


Ad-tech firms use cookie syncing to swap user IDs through redirects and pixels. See what Safari, Firefox and Chrome do about it and why analytics skips it.


Google caps GA4 calculated metrics at 5 per standard property. Learn the formula syntax, the units, where they appear and what a formula can't use.


Set up GA4 content grouping with the content_group parameter in gtag or Tag Manager, read the Content group dimension and keep (not set) rows out.


In GA4, Google Analytics annotations pin a dated note to line graphs. Who can add them, the 60 and 150 character limits, the 1,000 cap, and change log use.


A Google Analytics opt out can be a browser add-on, a ga-disable window flag or denied consent. What each one blocks, who controls it, and where it stops.


With Google tag gateway, the Google tag loads from your own domain and hits pass through your CDN. What it changes for data accuracy, and how to set it up.


The gtag function is how the Google tag takes config, event, set and consent commands. How gtag.js loads, and how it differs from Google Tag Manager.


In GA4, pages per session is called Views per session: page and screen views divided by sessions. The formula, where to find it, and how to read it.
Accurate SPA tracking means one pageview per screen with the right referrer. How the History API, virtual pageviews and GA4 history events fit together.


Chrome's user agent client hints split browser data into Sec-CH-UA headers. See low vs high entropy, Accept-CH, the reduced UA and what analytics reads.


Each band of a scroll map is colored by the share of visitors who reached it. Learn the fold line, the drop-off column, and where page height misleads.


Each conversion path is the ordered list of channels a buyer touched before converting. How GA4 splits it into early, mid and late touches, and what to fix.


Registering GA4 custom dimensions turns your event parameters into report fields. See the three scopes, the 50, 25 and 10 quotas and the cardinality trap.


Sending a GA4 user ID links one signed-in person's sessions across devices. See the gtag setup, the 256-character rule, sign-out handling and the limits.


GA4's average engagement time is focus time divided by active users, not sessions. How the timer runs, why the monthly figure is higher, and what it misses.
Website click tracking records which element a visitor clicked, on which page and when. See the three methods, the fields to send, and why some clicks vanish.


A data layer is the queue a page pushes analytics events into. See how Tag Manager reads dataLayer, why pushes merge, and the mistakes that drop events.


GA4 counts engaged sessions as visits that pass 10 seconds, fire a key event, or record two page views. How each trigger fires and where the count misleads.


Google caps GA4 event parameters at 25 per event. See the name and value limits, reserved names, and why custom ones stay out of reports.


Tracking micro conversions shows progress toward a sale or signup. The two types, how to pick them, and why they belong outside your main KPI.


Organic traffic is a classification rule, not a fact. See how GA4 and Matomo file a visit under search, and why the search keyword never arrives.


In analytics, user segmentation splits one report into groups that share a trait. See the segment types, how to build one, and when a segment is too small.


Google Ads basics: what is GCLID, how auto-tagging stores it, why redirects drop it and how it differs from UTM tags.


An active user means something different in every tool. Compare the exact active user definitions in GA4, Mixpanel, Amplitude, PostHog and Meta's 10-K.


Path analysis vs funnel analysis: a funnel measures steps you define, a path shows the sequences users took. How GA4, Mixpanel and Amplitude build each one.


Referer vs referrer: the HTTP header kept a 1996 misspelling, while document.referrer and Referrer-Policy spell it right. See which name goes where.


Session cookie vs persistent cookie from RFC 6265 and MDN: the Expires and Max-Age rule, session restore, the 400-day cap, and what EU consent rules say.


Session replay data retention compared from vendor docs: FullStory, Hotjar, Clarity, LogRocket, PostHog, Sentry and Datadog, plus what GDPR requires.


Session replay vs heatmaps: one visit against many, and how Hotjar, Datadog, PostHog, Clarity and FullStory build heatmaps from, or apart from, recordings.
A tracking plan lists every event, trigger, property and owner before code ships. See the columns, a filled example, and the gap validation tools miss.


A form abandonment rate compares forms started with forms submitted. See the formula, how GA4 and Zuko define a start, and a worked field-level example.


Legitimate interest GDPR rules in Article 6(1)(f), quoted from EUR-Lex, with the EDPB three-part test and why web analytics is still a contested case.


Local storage vs cookies compared from the specs: size limits, expiry, what travels with each request, HttpOnly, Secure, SameSite, Safari ITP and ePrivacy.


How rrweb records a session as a DOM snapshot plus mutations, not video: its npm packages, MIT licence, and privacy options like maskAllInputs and blockClass.


Schrems II, CJEU case C-311/18 of 16 July 2020: the operative part quoted, why Privacy Shield fell, what SCCs now require, and the 2023 Data Privacy Framework.


A session replay sampling rate sets what share of visits get recorded. See how Sentry, PostHog, LogRocket, Datadog and Clarity document it.


Time spent above the fold fell from 80% in 2010 to 57% in 2018 in NN/g eyetracking. Why a fixed pixel height is the wrong way to think about it.


In GA4, data sampling starts above 10 million events on standard properties and 100 million on Analytics 360, and one icon is the only on-screen tell.


A checkout funnel analysis covers only the steps from begin_checkout to purchase, which is where the leak is fixable and the event names are standard.


Every clickstream data row is one hit with a timestamp, an identifier and a URL, which is why the GDPR and the CCPA both treat the export as personal data.


The cookie size limit is 4096 bytes of name plus value in Chrome and Firefox, 1024 bytes per attribute, and 180 cookies per domain. With the source lines.


The entry page vs landing page split matters because GA4 ships only a Landing page dimension and Adobe ships only Entry page, computed from different fields.


An event naming convention has to satisfy GA4's reserved names, 40 character ceiling and case sensitivity, because renaming after launch never touches history.


A session replay GDPR review turns on four things: an Article 6 lawful basis, Article 9 special category data, ePrivacy Article 5(3) consent, and a DPIA.


The rage clicks vs dead clicks split is about cause, not severity. Detection thresholds published by PostHog, Hotjar, FullStory, LogRocket and Clarity.


Referral traffic is decided by two different rules. GA4 reads the medium, Matomo reads the referrer host, and the Referer header drops most of the URL.


SameSite Lax vs Strict vs None decided by the RFC text and Chrome's own documented default, including the Secure requirement and what each mode kills.


A sankey diagram conserves flow, so every unit entering a node has to leave it. Where that helps in analytics, where it misleads, and which tools render one.


A self reported attribution field catches podcasts, word of mouth and dark social that click tracking never logs. Why answers skew and how to reconcile them.


Vanity metrics are not a banned list of numbers. A metric is vanity when no plausible value of it would change what you do next, and that depends on you.
A tracking pixel is a 1x1 image request that reports a view. Here is what the server receives, what it cannot read without a script, and why opens overcount.


The PII definitions in NIST SP 800-122, GDPR Article 4(1) and the CCPA do not line up. Which identifiers each one names, and where they disagree.


PII redaction session replay defaults vary wildly: rrweb masks only passwords, Sentry masks all text, Clarity masks numbers and emails.


A channel with no matching rule, unassigned traffic ga4 shows is not a missing value like (not set). Google's own rules explain the difference.


The unique visitors meaning depends on the tool: GA4 estimates it, Matomo refuses to compute it on long ranges, Adobe dedupes it across the whole report.


Comparing user flow vs user journey separates one observed path through your product from the whole goal a person is chasing, across channels and over time.


Google defines utm_source vs utm_medium in one line each. GA4's channel rules decide which parameter actually sorts your traffic. Both, quoted exactly.


Google's docs give not set in GA4 a different cause per dimension, from a missing session_start to an empty content_group. Here is each cause and its fix.


Setting CHIPS partitioned cookies takes Secure, SameSite=None and one more attribute, and the browser keeps a separate copy per top-level site.


Reversibility decides pseudonymisation vs anonymisation. Article 4(5) data stays personal data. Recital 26 anonymous data leaves the GDPR for good.
Comparing sendBeacon vs fetch keepalive: both share one 64 KiB in-flight budget, but only sendBeacon locks your Content-Type and hides the server response.
Learn what server side tracking moves to your own server, which Safari cookie caps it escapes, how to deduplicate events, and why consent obligations stay.


The real answer to does session replay slow down website speed, with published rrweb and Sentry numbers for script size, main-thread CPU, and upload bandwidth.


Every session replay lawsuit wiretapping claim runs on CIPA 631, Pennsylvania's WESCA, or the federal Wiretap Act. What plaintiffs allege, what courts held.


Nine tracked industries return a documented average bounce rate by industry ranging from 35.76% to 48.38%, sourced from Databox data dated September 2024.


The average order value formula divides revenue by orders, and a single discount code or return policy can quietly distort every number a team reports.


Databox's own data puts average session duration benchmarks at 77.61 seconds for B2B sites and 92.33 seconds for B2C sites, split by industry and device too.


In classic analytics, average session duration gives zero recorded time to the very last pageview of every session, which quietly drags the average down.


Comparing beforeunload vs pagehide shows why mobile skips beforeunload, blocks the back-forward cache, and why pagehide with sendBeacon flushes data reliably.


Unlike a raw pageview count, behavioral analytics records what a visitor actually does on a page, as a stream of named events tied to that specific visitor.


A combination of canvas rendering, installed fonts, screen size and timezone is what browser fingerprinting turns into an identifier that survives deletion.


A bug report template names every field a reviewer expects, from steps to reproduce to severity, so incomplete tickets get rejected before an engineer looks.


A recurring meeting called bug triage classifies new reports by severity, assigns each one an owner, and keeps that process separate from backlog grooming.


Roughly seven in ten online carts never turn into an order, and the cart abandonment rate formula shows how much revenue that gap represents in a given month.


A retention curve only makes sense once cohort analysis groups users by a shared start date, since one average hides the pattern a cohort table reveals.


Step conversion and overall conversion answer different questions about a conversion funnel, and the gap between them shows exactly where drop-off happens.


The CPRA amended the CCPA in 2023, adding sensitive personal information rules, a right to correct data, and a dedicated enforcement agency, the CPPA.


The customer acquisition cost formula divides total sales and marketing spend by new customers won, and the numerator decides if that number means anything.


Unlike a journey map drawn in a workshop, customer journey analytics measures stitched sequences of real behavior across channels, sessions and devices.


Every DAU/MAU ratio benchmark in circulation traces back to three sources: Fred Wilson in 2011, an undated Gainsight guide, and Mixpanel's June 2026 report.


Unlike event counting, digital experience analytics reconstructs the visit itself, using session replay, heatmaps, friction detection and journey analysis.


Seeing direct / none in GA4 means a session arrived with no referrer and no campaign tag. How it differs from (not set), Unassigned and the Direct channel.


Sessions land in direct traffic when no referrer and no campaign tag survive the hop. The real causes: referrer policies, apps, PDFs, redirects, QR codes.


California puts the Do Not Sell or Share My Personal Information link in a homepage header or footer, with the exact title fixed by Civil Code 1798.135.


Every funnel produces two drop-off rate numbers, one per step and one end to end, and teams quote them interchangeably. A worked table separates them.


Search engines own dwell time and your analytics cannot see it. Where the line falls against time on page and session duration, and what Google documents.
In product analytics, event tracking records named user actions with properties attached, which turns page-level traffic into a record of what people did.


Three setup choices decide what funnel analysis reports: step sequencing, the conversion window, and whether the funnel counts users or sessions.


Most published funnel drop-off benchmarks average companies that define steps differently, so the numbers rarely transfer. Every figure here is sourced.


Google applies GA4 data thresholding to withhold demographic rows from a report. What triggers it, why the cutoff is undocumented, and what reduces it.


What Google Consent Mode v2 added: ad_user_data and ad_personalization, the basic versus advanced choice, and what stops working in Google Ads without them.


Under the GDPR, is an IP address personal data whenever the site operator has legal means to identify the visitor, the CJEU held in Breyer. US law is split.


Google split one word in two: key events vs conversions in GA4 now name different objects in Analytics and Ads. What each one means and where each shows up.


One net revenue retention formula, five defensible variants: the same cohort returns 84.0%, 104.5%, 108.3%, 109.5% or 110.3% depending on window and base.


Before launching a test, a/b test sample size determines whether the result you see is signal or noise, and the formula needs three numbers set in advance.


Move the attribution window from 7 to 90 days and one $1,800 order pays three different sets of channels. See the arithmetic and the current platform defaults.


In product analytics, autocapture records every click, page view and form submit automatically, without a single tracking call written by hand.


Reading bounce rate vs exit rate comes down to two denominators. One pricing page returns a 40% exit rate and a 15% bounce rate in the same week.


One churn rate formula, three defensible denominators: the same month of data returns 3.70%, 4.00% and 6.00% depending on what you agree to divide by.


The conversion rate formula is conversions divided by sessions or unique visitors, times 100. One week of traffic run through both denominators gives two.


A cookieless analytics tool counts visitors without storing an identifier in the browser. What that removes, what it costs, and why fingerprinting fails.


Every Cumulative Layout Shift score is impact fraction times distance fraction, and Google reports the largest burst of shifts, not the sum of them.


The GDPR test for data controller vs data processor is who determines the purposes and means. What each role signs, owes, and does when a breach hits.


GA4 reports engagement rate as engaged sessions divided by total sessions. A session counts as engaged at 10 seconds, one key event, or two page views.


The split between first-party vs third-party cookies is the domain that set them, not who wrote them. What browsers block now, and what breaks when they do.


The four frustration signals are rage clicks, dead clicks, error clicks and thrashed cursors. Each one fires on a threshold your tool sets, not on a guess.


Chrome scores Interaction to Next Paint on a page's worst interaction, timing input delay, processing and presentation delay to the next painted frame.


Scored good at 2.5 seconds or less, Largest Contentful Paint splits into four subparts, and the split is what tells you which one is costing you the seconds.


Credit for one conversion depends on the marketing attribution model you pick. See seven models split the same four-touch journey, with exact percentages.


The MRR meaning covers new, expansion, contraction and churned revenue each month, and excludes one-time fees that never repeat.


Comparing pageviews vs sessions vs users shows three separate counts that roll up into each other and rarely land on the same number twice.


Unlike web analytics, product analytics measures behavior event by event and user by user, so you can answer activation, retention and funnel questions.


Passive by design, real user monitoring collects performance and error data from visits that really happened, then reports it at p75 instead of an average.


The retention rate formula divides retained customers by the customers you started with, and miscounting new signups skews the result every time.


Apple's Safari 7 day cookie limit caps script-set first-party cookies, and a shorter 24 hour cap applies after certain cross-site link clicks.


A session timeout closes an idle visitor's session, and the default 30-minute rule is why the same traffic reports different counts.


A good bug report writes steps to reproduce as numbered actions from a known starting point, so anyone on the team can trigger the same failure.


A plain definition of web analytics, what a tracking script collects, the core metrics and how each one gets misread, and where product analytics takes over.


The phrase website heatmap covers five overlays: click, scroll, move, confetti and attention. Here is what each one measures and where each one misleads.


Benchmarks answer what is a good conversion rate differently for ecommerce, SaaS and lead gen, and a single median number hides more than it reveals.


In web analytics, a session is one group of a visitor's interactions, closed by inactivity, midnight, or a campaign change.


A session replay rebuilds a visit from DOM mutations and input events, not video. See what it captures, what masking hides, and how it differs from heatmaps.