Web analytics Study Guide

📖 Core Concepts Web Analytics – measurement, collection, analysis, and reporting of web data to understand and improve website performance. On‑Site vs Off‑Site Analytics – on‑site tracks behavior after a visitor lands on your site; off‑site gauges audience size, share‑of‑voice, and buzz outside your domain. Key Performance Indicator (KPI) – a metric tied to business goals (profit, cost‑savings, market‑share) used to judge success. Session (Visit) – continuous activity from a single visitor; ends after 30 min of inactivity (platform‑specific). Unique Visitor – a distinct client identified (usually via cookies) during a reporting period. --- 📌 Must Remember Bounce Rate = % of visits with only one page view. Exit Rate = % of visits where a given page is the last page viewed. Average Page Depth = total page views ÷ total visits. Frequency = total sessions ÷ total unique visitors. Session Duration = total session time ÷ total sessions. Click‑Through Rate (CTR) = clicks ÷ impressions. Hotel Problem – daily unique‑visitor counts ≠ monthly unique visitors because repeat visitors are counted multiple times. Analytics Poisoning – bot traffic that inflates metrics. First‑Party vs Third‑Party Cookies – first‑party set on your domain, less likely to be blocked or deleted; third‑party can track across sites but are often blocked. --- 🔄 Key Processes Data Collection – capture raw events (page requests, clicks) via server logs or page‑tag JavaScript beacons. Processing → Metrics – convert counts into ratios (e.g., bounce rate = bounces ÷ visits). KPI Development – align metrics with business objectives (e.g., target CTR > 5 %). Online Strategy Formulation – set goals, objectives, standards (profit, cost‑savings, market share). A/B Testing – run two variants, measure a statistically‑tested result, adopt the higher‑performing version. --- 🔍 Key Comparisons On‑Site vs Off‑Site Analytics On‑Site: measures visitor behavior inside the site (page views, click path, conversions). Off‑Site: measures potential audience, buzz, and share‑of‑voice across the web (keyword identification, open‑data analysis). Cookies vs IP + User‑Agent Cookies: reliable per‑visitor ID, but can be blocked/deleted. IP + User‑Agent: improves identification when cookies unavailable but still ambiguous (shared IPs, identical agents). First‑Party vs Third‑Party Cookies First‑Party: set on your domain → higher retention, less privacy pushback. Third‑Party: track across domains → higher coverage but more likely to be blocked. Log‑File Analysis vs Page Tagging Log‑File: server‑side, captures every request, misses client‑side events (scrolls, hovers). Page Tagging: client‑side JavaScript, captures rich interactions, dependent on browser JS execution. --- ⚠️ Common Misunderstandings “Unique visitors = total visits” – they are different; a repeat visitor contributes multiple visits but counts once as a unique visitor. “Higher bounce rate always bad” – on a single‑page site or blog post, a high bounce may be normal. “Cookies give 100 % accurate IDs” – cookies can be blocked, deleted, or shared across devices, leading to over‑ or under‑counting. “All bot traffic is harmless” – bots cause analytics poisoning, inflating traffic and skewing conversion rates. --- 🧠 Mental Models / Intuition “Traffic Funnel” – imagine visitors moving through stages: Impressions → Clicks (CTR) → Page Views → Sessions → Conversions. Each stage’s metric is a ratio of the previous stage. “Session as a TV Show Episode” – a session starts when the viewer presses play (first request) and ends after a commercial break (30 min inactivity). “Cookie as a Loyalty Card” – it sticks to a user across visits, but if the card is lost (deleted) the store treats them as a new guest. --- 🚩 Exceptions & Edge Cases Session Timeout Variability – different platforms use 30 min, 20 min, or custom thresholds. Multi‑Device Visitors – same person on phone and desktop gets counted as separate unique visitors if cookies differ. Geolocation Accuracy – IP‑based location can be wrong for VPNs, mobile carriers, or proxy servers. Impression vs View – an ad impression may be counted even if the page is never fully rendered (e.g., lazy‑loaded images). --- 📍 When to Use Which Choose Log‑File Analysis when you need a complete server‑side picture and cannot rely on client‑side scripts (e.g., high‑security environments). Choose Page Tagging for rich interaction data (scroll depth, clicks, form submissions). Use First‑Party Cookies when privacy regulations are strict or you need consistent visitor IDs. Apply A/B Testing for any change that impacts a measurable KPI (CTR, conversion rate). Select Off‑Site Keyword Identification when planning SEO or paid search campaigns; use on‑site click path analysis for UX improvements. --- 👀 Patterns to Recognize High Bounce + Low Exit on a Page → likely a landing page that satisfies the visitor’s need in one view. Spike in Sessions but Flat Unique Visitors → possible bot traffic or analytics poisoning. CTR ↑ while Impressions ↓ → targeting is becoming more precise (or ad inventory is shrinking). Session Duration ↑ + Conversion ↓ → users are spending more time but not converting – possibly usability issues. --- 🗂️ Exam Traps Confusing Bounce Rate with Exit Rate – bounce = single‑page visit; exit = last page of any multi‑page visit. Assuming “Total Visits = Unique Visitors” – repeat visitors inflate total visits. Treating All Cookies as First‑Party – many platforms still rely on third‑party cookies; the exam may ask which metric is affected when third‑party cookies are blocked. Ignoring the Hotel Problem – a monthly unique‑visitor total cannot be obtained by simply summing daily uniques. Mixing Up CTR and Conversion Rate – CTR = clicks ÷ impressions; conversion rate = conversions ÷ clicks (or visits). ---