Analytics & measurement
Privacy-friendly analytics, events, funnels, A/B tests, KPIs and measurement errors.
A/B testing pitfalls: sample mismatch, novelty and peeking
Sample ratio mismatch, repeated peeking, novelty effects and after-the-fact segments: the practical errors that distort A/B tests and how to spot them.
A/B testing a website: designing the test before launching it
Hypothesis, primary metric, sample size and duration: how to design an A/B test on a website so that the result is reliable and genuinely useful.
Bot and spam traffic in analytics: spotting and filtering it
How to recognise bots, crawlers and referral spam in analytics data, which signals to watch and how to filter them without losing real visits.
Bounce rate and time on page: what they actually measure
How bounce rate, time on page and engaged sessions are calculated, why they mislead and which signals to use to judge whether a piece of content is working.
Cohort analysis: measuring retention without being misled
What a cohort is, how to read a retention table, why overall averages mislead and how to use cohorts to evaluate a website or online service honestly.
Attribution models: who gets the credit for a conversion
Last click, first click, linear, time decay and position-based: how attribution models work, what each one rewards and the limits they all share in practice.
Conversion funnels: how to build one and read it properly
How to define funnel steps, choose between open and closed funnels, calculate step conversion rates and find the exact point where visitors drop away.
Cookieless analytics: measuring a website while respecting privacy
How cookieless analytics work, what you give up compared with classic tracking, and how to choose a lightweight tool that treats visitor data with care.
Cross-domain and subdomain tracking: sessions without breaks
Why moving between domains, subdomains and payment services breaks analytics sessions, and how to set up consistent cross-domain measurement.
Data anomalies: investigating a sudden drop in traffic
How to investigate a sudden drop or spike in analytics: telling measurement errors from real change, breaking data down and finding the cause.
Data layer: structuring it well and keeping it healthy
What a data layer is, why it separates the site from measurement tools, how to structure context and events, and how to stop it decaying over time.
Analytics dashboards: designing one people actually read
How to design a useful analytics dashboard: a defined audience, a few questions, meaningful comparisons, suitable charts and upkeep that keeps it alive.
Essential e-commerce metrics: conversion, order value, worth
The key metrics for an online shop: conversion rate, average order value, revenue per visit, repeat purchase, returns and margin, read together.
Event taxonomy: names and properties that stay readable
How to give analytics events consistent names, pick the right properties and document everything in a dictionary, so reports stay clear for years to come.
Form analytics: understanding where and why people give up
How to measure a form field by field: starts, validation errors, blocking fields, timings and abandonment, without recording what users type.
Heatmaps and session recordings: when they are worth using
What heatmaps and session recordings really show, when they help, how to read them without being misled and how to respect visitors' privacy.
Incrementality: measuring real impact with holdout groups
What incrementality is, why attributed conversions are not enough and how to estimate it with holdout groups, geographic tests and planned pauses.
Website KPIs: vanity metrics versus actionable metrics
How to tell vanity metrics from those that drive decisions, choose a handful of KPIs for a website and pair them with guardrail and context metrics that help.
Measuring content reading: scroll, visibility and attention
How to tell whether articles are really read: scroll depth, section visibility, active time, follow-up clicks, and the limits of each measure.
Server logs or JavaScript analytics: two measurements compared
Why server logs and JavaScript analytics report different numbers, what each method actually sees and how to combine them for a credible picture of traffic.
Server-side tracking: how it works and when it pays off
What server-side tracking is, how it differs from in-browser tracking, what it offers for data control and the costs and duties it brings with it.
Site search analytics: what search terms teach you
How to analyse a website's internal search: queries, zero-result searches, refinements and exits, to improve the content and the site navigation.
Checking tracking before release: a practical QA procedure
A procedure for checking events and conversions before every release: test environments, test cases, browser tools, automation and alerting.
UTM parameters: tracking campaigns without creating confusion
What the five UTM parameters are, how to fill them in with consistent rules, which mistakes pollute your reports and why they never belong on internal links.
Measurement plan: deciding what to track before tracking it
How to write a measurement plan for a website: goals, questions, metrics, events and ownership, with a concrete example and the most common mistakes to avoid.