Web analytics
Web analytics, explained simply — what the numbers mean, how to set up tracking that answers real questions, and how analytics works on the platforms you build with.
Server-side vs client-side tracking, explained without jargon
Almost every analytics tool works the same way: a script in the visitor's browser reports back what happened. Server-side tracking moves that reporting to a computer you control — which is why ad blockers can't stop it, and why it costs real money to set up. Here's how both work, what switching actually recovers, and when it's worth it.
Cookieless analytics: how counting without cookies actually works
'Cookieless' is a single label covering at least three different techniques, and they trade away different things — mostly the ability to recognize a visitor who comes back tomorrow. Here's how each approach counts people, why cookieless tools often show more traffic than Google Analytics, and where the no-banner-needed claims actually stand in 2026.
Session replay lawsuits, explained for site owners
Thousands of websites have received wiretapping lawsuits or demand letters for running session replay and tracking pixels — under a California law written in 1967. Most of them could have avoided it with one change: nothing records until the visitor clicks Accept. Here's what the lawsuits actually claim, what the courts have decided so far, and what to check on your own site.
Dark traffic: the visitors your analytics can't see
Ad blockers stop the analytics script for roughly a third of the web. Privacy browsers strip referrers and cap cookies. Consent banners never get an answer from half of desktop visitors. And most private shares arrive with no referrer at all, landing in Direct. Your dashboard shows a sample, not the universe.
Why your analytics numbers never match: GA4 vs ads platforms
Your Google Ads campaign reports 150 conversions. GA4 reports 80. The gap exists because each system answers a different question and counts conversions by different rules — attribution windows, view-through credit, modeled conversions, click dates. Here's why, and which number to use for what.
Identify website visitors: from anonymous visit to customer profile
Strangers can't be unmasked — legitimate analytics doesn't identify people who haven't identified themselves. But your own users are different: choose to recognize them at signup, and their entire anonymous pre-signup history merges into a named profile. Here's what that makes possible.
What's a good conversion rate? Benchmarks by industry and traffic source
Around 2% for an online store, 3–6% for a lead-gen site, and 6%+ for a dedicated landing page — the 'average conversion rate' depends entirely on what's being counted. Here are the 2026 numbers from sources that publish their methodology, and how to read them without fooling yourself.
What counts as a “session”? Every tool defines it differently
Most analytics tools end a session after 30 minutes of inactivity — but GA4 stopped splitting sessions at midnight, Matomo still does, Fathom uses a 24-hour window, and some privacy tools don't count sessions at all. Same traffic, different numbers, and almost nobody discloses which rules they're playing by.
Average session duration is lying to you
The last page of every visit contributes zero seconds, bounces drag the average toward nothing, and whether the clock keeps running while a tab sits in the background depends on which tool you ask. Here's how the number is actually computed, why it fell when you switched to GA4, and what to look at instead.
What is web analytics? A plain-English guide
You add one small script to your site and it starts counting: how many people visited, which search or link brought them, and whether any of them signed up. This guide explains how that works, what the terms on the dashboard mean, and how much to trust the numbers.
Why did my traffic drop? A step-by-step diagnosis
Most traffic drops have one of five causes: broken tracking, seasonality, a channel-specific problem, a Google update, or AI Overviews eating your clicks. Check them in that order — the common causes are also the fastest to rule out.
How to track website conversions
A conversion is just the action you want a visitor to take. Which action it is, is your call — and the choice decides everything the number is worth.
How to check website traffic — for your site or any site
Checking your own site takes one snippet and two minutes for exact numbers. Checking someone else's takes an estimator and a grain of salt. Both routes, free — and the accuracy data the tool pages don't mention.
How accurate are website traffic checkers?
Screaming Frog compared estimators against real analytics on 25 sites: Semrush ran about 42% low, Ahrefs 36%, and SimilarWeb averaged closest with the wildest variance. Where the numbers come from, and when to trust them.
Why do different analytics tools show different numbers?
Run two analytics tools on one site and they'll disagree by 10–30% — with both correct by their own rules. The six causes of the gap, which number sits closest to the truth, and how to work with it.
What's a good amount of traffic for a website?
A consultancy can thrive on 300 visits a month while a content site starves on 50,000. The benchmark worth having comes from your own numbers: monthly goal divided by conversion rate.
Who is visiting my website? What you can (and can't) see
No tool will hand you the names of anonymous visitors — anyone promising that is selling trouble. What you can see: sources, journeys, company names on B2B traffic, and everything, once someone signs up.
I just launched my product. What should I track first?
Launch week needs five numbers, and most analytics tools ship with forty. Visitors, sources, one action, signups and one activation signal — in that order. Everything else can wait for month two.
The only 5 metrics to check every morning
A morning check should take three minutes, not a dashboard safari. Five numbers with yesterday's baseline attached — and the popular ones that deliberately aren't on the list.
Session replay, explained (with examples)
A session recording is a reconstruction, not a video — a distinction that decides what it can show, what it must hide, and why it costs so little to run. How replay works, three real failure patterns, and the workflow that keeps it from eating afternoons.
Heatmaps vs. session replay: what's the difference?
One shows a thousand visitors on a single page, the other shows one visitor across a whole visit. When each is the right tool — and the specific ways each one lies.
Best Hotjar alternatives (2026)
Hotjar is being folded into Contentsquare, and a lot of teams are shopping. The alternatives compared in plain language — what each one is for, and where the catches are.
What events to track on your website
There's one event worth more than all the rest: the action that predicts whether a new user sticks around. Here's how to find it — and the short list of others actually worth tracking.
Traffic sources: direct vs. organic vs. referral vs. paid
Four buckets decide where your visitors 'came from' — and the biggest one is a catch-all for traffic the tool couldn't identify. What each source really means, why 'direct' misleads, and why the top source of visits is rarely the top source of revenue.
Nobody signed up. Here's what to check first
Zero sign-ups after a launch rarely means the idea flopped — usually it's a broken form, an invisible offer, or a vague pitch. The order to check them in, before you touch the product.
How to see where users get stuck on your website
Session recordings and heatmaps turn a drop-off number into a video of the exact moment someone gave up. The four friction patterns to look for, and the four-step workflow for finding yours.
A simple Google Analytics alternative that shows revenue
GA4 can answer almost any question — after a property, a data stream, a tag manager, and a weekend. Here's what a simpler tool needs to cover, and where the trade actually sits.
Checkout abandonment rate: benchmarks and how to reduce it
Roughly seven in ten started checkouts never finish. What counts as normal, how to find your leak, and which fixes actually move the number.
What is a good bounce rate?
Under 40% is good, over 70% is a flag — except for blogs, landing pages, and every other case where it isn't. What the number means and when to ignore it.
UTM parameters: a practical guide
Five little tags decide whether 'is my marketing working?' has an answer. How UTMs work, the naming rules that keep data clean, and the one place never to use them.
Page views vs. visits vs. unique visitors
One person, two coffee breaks and seven pages is 1 visitor, 2 visits and 7 pageviews. Which number to quote, when, and why no two tools ever agree.
Lovable's built-in analytics: what it tracks and where it stops
Lovable ships a traffic dashboard with every published app. It answers 'did people come?' well — and goes quiet the moment you ask what they did.
Replit's built-in analytics: what it tracks and where it stops
Replit ships genuinely good infrastructure telemetry with every published app — requests, uptime, CPU. The catch: a request isn't a person, and none of it says what users actually did.