Analytics and Tracking

Web Analytics

Also called site analytics, digital analytics

The practice of recording what visitors do on a website and reading it as behaviour you can act on.

Quick facts: Web Analytics

Category
Analytics and Tracking
Also called
site analytics, digital analytics
Level
Beginner
Affects
Marketing decisions, budget allocation, website improvements
Where to see it
GA4, Google Search Console, Looker Studio, server or hosting logs
In this article4
  1. What web analytics measures
  2. Why web analytics matters
  3. Common mistakes with web analytics
  4. How to act on it

What web analytics measures

Every time someone opens a page on your site, the analytics tool writes down a small bundle of facts: which page, which device, which country, where the visit came from, and what happened next. Web analytics is the practice of collecting those records and reading them as behaviour rather than as raw hits. It covers the website only — not what an ad platform reports about its own clicks, and not what your sales system knows about the same person once they become a customer.

Modern tools such as GA4 build the picture from events. A page opening is an event, a scroll is an event, a form submission is an event. Each event carries parameters describing it, and the report you eventually read is those events counted and grouped. That chain is worth understanding, because a report can only ever show what was captured in the first place.

Why web analytics matters

Without it you are guessing which part of your marketing works. Web analytics is what connects money spent to behaviour observed: it shows that visitors from one channel read several pages and enquire, while visitors from another leave from the page they landed on. That is the difference between renewing a channel and cutting it.

In Nepal it earns its keep for a second reason. Traffic here is heavily mobile and often on slow or intermittent connections, so a page that behaves well on a laptop can quietly fail for most of the audience. Analytics shows that split long before it shows up as lost enquiries.

Common mistakes with web analytics

The most common is treating the numbers as exact. Ad blockers, consent choices, browser privacy settings and offline conversions all remove real behaviour from the record, so two tools reporting the same thing will rarely agree. Analytics is a reliable guide to direction and proportion. It is not an accounting ledger, and it should never be used as one.

The second is collecting everything and deciding later. An account with hundreds of events and no stated question produces reports nobody opens. Start from a measurement plan that names the decisions the data has to support, then track that.

The third is comparing periods without recording what changed. A drop that looks alarming is very often a tracking change, a site migration or a season, and nobody can tell which if no one wrote it down.

How to act on it

Pick a small set of outcomes that represent real business value — an enquiry, a booking, a purchase — and make sure those are recorded properly before you worry about anything else. Everything further up the funnel is context for them.

Then read the data by dimension rather than in total. Totals hide the story; the same conversion rate split by device, channel and landing page usually points straight at the fix. And if the collection layer itself is broken, correcting it comes first — no amount of reporting rescues data that was never captured. That correction belongs in the analytics and tracking setup, not in the dashboard.

Do and do not

Do

  • Record real business outcomes before anything else
  • Read every number split by device and channel
  • Write down what changed, so drops can be explained

Do not

  • Treat analytics figures as an accounting ledger
  • Collect events with no question behind them
  • Compare periods without checking for tracking changes

Questions people ask about this

Is web analytics the same as Google Analytics?

No. Google Analytics is one tool for doing web analytics, and it is the most widely used one, but the practice is bigger than any single product. Server logs, ecommerce platform reports, heatmap tools and your own customer database all answer behavioural questions. The skill is choosing the right source for the question and knowing what each source cannot see.

Why do my analytics numbers not match my ad platform?

Because they count different things. An ad platform credits a conversion to the click it served, inside its own attribution window and sometimes using modelled data. Your analytics tool credits the last session it could observe, and it loses visitors who block scripts or decline cookies. Expect a gap, watch the trend rather than the absolute figure, and pick one source as the decision-maker.

How much traffic do I need before analytics is useful?

Less than people assume for diagnosis, more than people assume for testing. Even a quiet site shows you which pages visitors leave from and which channels bring enquiries. Comparing two versions of a page reliably is a different job: that needs enough conversions to separate a real difference from noise, which small sites rarely reach quickly.

Related terms

Found this useful?

Share it, or ask an AI to summarise it

Back to the glossary

Knowing the term is the easy part

Applying it to your own site and budget is the work. Book a call and I will tell you what actually applies to you.