Web analytics is the collection, measurement and reporting of website data, which is used to understand how people find and use a site. As a B2B marketer, you can use website analytics to decide which channels to fund, assess whether a campaign earned its budget, and understand whether or not you can optimise your site for better conversions.

This guide covers what the term "website analytics" means, how the data is produced, what it measures, and where it has limitations that require additional tools.

And that last part matters more than most guides acknowledge, as web analytics counts browsers (not people), and it can't tell you which companies were behind the visits. If you don't understand where it falls short, it's easy to report numbers that are accidentally misleading.

What is web analytics?

Web analytics collects data from a website, processes it into structured reports, and presents those reports so that someone can act on them. At its simplest, it answers four questions:

• How many visits the site got

• What traffic sources they came from

• What they did once they landed on your site (including if they took any actions or what pages they visited)

• Whether they took specific conversion actions you're optimising for, like signing up for a newsletter or making a purchase.

Why one buyer shows up as three users

Consider one buyer evaluating your product over a week-long period.

She visits your site on Monday from her work laptop, and views three pages. On Wednesday she checks back from her phone and reads four more. On Friday, working from home, she returns on her personal laptop and looks at five pages.

Your analytics report records three different sessions, twelve pageviews and, depending on the tool's identifier logic, either two or three users. It doesn't record one person, so you're operating off incomplete data.

If she clears her cookies between visits, the tool might create a fresh identifier, which can cause the overcount to increase. If a colleague at the same company visits from the same office network, the tool logs another new user, not a second person at the same account. The data says nothing about the company driving the users that are visiting.

This isn't a configuration problem you can fix, because it's a structural feature of how the technology currently works.

The unit of measurement is the browser session, and a browser session is not the same thing as a single buyer. Metrics downstream (including unique users and conversion rates) all inherit that limitation.

And for a B2B marketer reporting monthly traffic, it matters. The 8,000 users in last month's report weren't actually 8,000 people, and you know that, but the report can't tell you how many they were.

Critically, none of this makes the data useless. Traffic trends, channel comparisons, content engagement and goal completions are all directionally reliable when measured consistently over time.

The mistake, though, is treating a user count as a headcount, or expecting a tool that counts browsers to answer a question about the companies that could become customers.

Web analytics, website analytics, site analytics: the same thing

You'll see "web analytics" used interchangeably with terms like "website analytics," "site analytics," and "analytics for websites." They refer to the exact same thing.

There's no technical or practical distinction between them, and if you searched for any of these terms and landed here, good news: you're in the right place!

This guide uses "web analytics" throughout because it's the most common phrasing, but nothing changes if your team or your client calls it website analytics or site analytics. The tools, data and key limitations are all the same.

How web analytics works

The data that ends up in your report passes through a chain of stages. At each stage, here's what happens:

1. The tag loads. When a visitor opens a page, the JavaScript snippet in the page code executes and begins listening for interactions.

2. An event fires. Each interaction, a pageview, a click, a scroll or a form submission, generates an event that the tag sends to the analytics tool.

3. An identifier is assigned. The tool checks for an existing cookie in the visitor's browser. If one exists, the event is appended to that visitor's session. If not, a new identifier is created.

4. Events are aggregated. Over the reporting window, the tool groups events by identifier, session and time, applying filters for bots and internal traffic along the way.

5. The report renders. The aggregated data appears as charts, tables and dashboards, ready for review.

The collection method shapes what you can see.

Client-side tagging is what most B2B sites use because it's the easiest to install and gives the richest behavioural detail. But it has a blind spot: any visit where the JavaScript doesn't fire, whether because of an ad blocker, a declined consent banner or a browser restriction, never reaches the report.

Server-side tagging closes some of that gap by moving the first collection point to your own infrastructure, though it requires more engineering work.

Server log files capture everything the web server touched, but without the client-side context that makes the data useful, and with heavy noise from bots and crawlers that the other two methods would have filtered out.

Here's a breakdown of the three options and the differences between them:

As a result, two tools measuring the same site won't produce the same numbers. One may use client-side tagging while the other reads server logs, or one may define a session as 30 minutes of inactivity while another uses 20. Neither is "wrong," and this is fairly normal across reporting tools, but it does make it difficult for teams to track if they don't understand.

What matters is keeping your own decisions consistent over time so the trends hold up.

For how to do that well, see our guide to web analytics best practices. For the practical steps of setting up tracking itself, start with how to monitor website traffic.

Why web analytics matters

Web analytics can help you determine:

Where to spend. When your acquisition report shows that paid search converts at twice the rate of display, the budget conversation starts with evidence rather than opinion.

What to fix. Pages with high traffic and low engagement are visible in the behaviour data, and so are the exit points where visitors leave before converting.

Whether a campaign worked. A landing page with 400 sessions and two demo requests tells a different story from one with 200 sessions and fifteen.

Where the channel is organic search, it has its own discipline: seo web analytics, with its own seo website metrics and seo reporting.

Those numbers, alongside every other channel, end up in the monthly report. For how to put that document together, see our guide to writing a web analytics report.

What web analytics measures

A website can count three things: visits (sessions), people (users) and companies. Standard web analytics tools count the first two and leave out the third.

Sessions are what most reports count. A session starts when someone arrives on your site and ends after a period of inactivity, typically 30 minutes. It captures what happened in a single sitting: pages viewed, buttons clicked, forms submitted, files downloaded.

Users are the tool's estimate of distinct individuals. It assigns an identifier, usually a cookie, and treats each identifier as one user. As covered above, one person across two devices becomes two users, and a returning visitor who cleared her cookies becomes a new one. The number overcounts in some directions and undercounts in others.

Companies are where B2B buying happens, and standard web analytics doesn't count them at all. Nothing in a typical analytics report tells you that the session on your pricing page came from a 200-person manufacturer in your target market.

For example, a monthly report might show 1,200 sessions, 900 users, from 140 companies. These are illustrative numbers, but they show the gap: the first two come from your analytics dashboard. The third doesn't.

Those three units sit underneath four families of reports:

These four families answer different questions, and they get confused with each other more often than you might expect.

The most common mistake is reporting audience and acquisition numbers every month, users up, organic traffic up, while never connecting them to the conversion column. When this happens, the traffic report looks healthy but the pipeline doesn't move. You only see the problem when you read the second column and the fourth together.

Want to go deeper on any of these metrics? Start here:

• Web analytics glossary - definitions for every metric in the table above

• B2B website metrics - the metrics that matter most for pipeline

• Website metrics - the broader list of what's worth tracking

• Web analytics dashboard - how to put it all in one view you check weekly

How to start with web analytics

If you're setting up analytics for the first time, or starting over after inheriting a setup nobody maintained, here's how to get to a useful baseline.

1. Decide what question you need analytics to answer

Write down the one question you want your analytics to answer. Two real examples: "which channels bring visitors who request a demo" and "which pages do target-account visitors read before they convert". The failure mode is deciding to track everything and reporting nothing. Once the question is on paper, set the targets against it: our guide to website KPIs covers how to set targets that actually mean something.

2. Choose an analytics tool

Before you pick a product, know what you need from it. The table below covers what to evaluate before you commit.

The last row is the one most B2B teams overlook. Standard analytics counts sessions and users. If your question from step 1 is about companies, you need a tool, or a layer on top of your existing one, that can report at that level. If you run analytics for clients, this is also the row that answers "is the traffic you send us the right traffic?" Without company-level data, you can show volume, but not whether the visits came from companies the client actually sells to.

Google Analytics 4 is free, and it is where most B2B teams start. If you want to compare alternatives, see our roundup of web analytics tools, and for the thinking that should come before any tool choice, see web analytics strategy.

3. Install and verify the tracking

Every analytics tool gives you a JavaScript snippet to paste into your site's code.

Once it's live, open your analytics tool's real-time view, visit your site in a separate tab, and confirm that the visit shows up within a minute. If it doesn't, the tag isn't firing. Check that it's placed before the closing </head> tag and that no consent banner is blocking it from loading. Most broken setups were broken from day one and nobody checked.

4. Set a reporting cadence

Decide how often you'll look at the data and what you'll do with it.

A simple starting point:

• Check acquisition and conversion numbers every week so you catch drops early

• Run a monthly review that answers the question you wrote down in step 1

Put both in the calendar. If nobody's opening the dashboard between meetings, the cadence falls through the cracks.

Web analytics by topic

This guide covers the fundamentals. These go deeper on specific problems.

Web analytics and website visitor analytics

Your analytics dashboard shows totals. You might see 1,200 sessions, with 40% coming from organic and a 15% bounce rate on the pricing page. That's useful for spotting trends, but it can't show you what any single visit actually did.

Visitor analytics can. It replays each session individually, showing which pages were viewed, in what order, how long the visitor stayed on each one, and where they dropped off. You can watch one visit move from a blog post to the pricing page to a case study in a single sitting.

If your pricing page has a 60% exit rate, the aggregate number tells you people are leaving. Visitor analytics shows you why. Maybe they're scrolling past the feature table and leaving at the FAQ, or they're bouncing before the page even loads on mobile.

The aggregate view answers the questions that matter most of the time, such as which channels work, where visitors drop off, and how many convert.

The visitor-level view is most useful when you need to understand behaviour the averages obscure, such as how high-intent visitors navigate compared to casual browsers, or whether the same visitor came back to your pricing page three times in a week.

If you need analysis that goes deeper than either view, such as segmentation, cohort comparison and multi-touch paths, our guide to advanced website analytics covers what comes next. Neither view, though, tells you which company was behind the visit.

What web analytics cannot tell you

Web analytics answers how many, and cannot answer who.

Web analytics tells you what happened on your site. It doesn't tell you what happened next, whether the visitor entered your pipeline, became a customer, or never came back. That connection lives in your CRM, and bridging the two is a critical part of B2B marketing analytics.

It also can't tell you who visited. Your report might show 8,000 sessions last month, but it won't show that 140 of them came from companies in your target market. Those are two different things, and mixing them up is where most reporting goes wrong.

A Gartner survey (March 2026) found that 67% of B2B buyers prefer a rep-free buying experience, and Leadfeeder product data shows that only about 2% of website visitors ever fill out a form. Your analytics tool recorded those sessions, but it can't tell you which companies they belonged to.

Browsers count, people do not

Some things make your user count too high. Others make it too low.

• Safari deletes JavaScript-created cookies after seven days without a visit (WebKit, April 2023). Cookies set by the server behave differently. Safari held 15.84% of global browser share in August 2026 (StatCounter), so on any site with Safari visitors, the user count runs high.

• One buyer using a work laptop, a phone and a home machine can show up as two or three users.

• When a visitor declines cookies or runs an ad blocker, the analytics tag never fires and the visit is never recorded. This hits EU traffic and technical audiences hardest.

The number overcounts browsers and undercounts people, and the two errors do not cancel out. If most of your traffic comes from Safari, your user count is probably inflated. If most of it comes from the EU with strict consent settings, you're probably missing visits entirely. Figure out which one hits your site harder, as that tells you more than trying to get the number exactly right. If you want to count companies rather than browsers, start with company level identification.

Here's a quick breakdown of the different factors to keep in mind:

From counting visitors to identifying companies

This is where company-level identification comes in. It is where web analytics meets the broader practice of website intelligence: understanding not just what happened on your site, but which organisations were behind it.

Company-level identification works by matching a visitor's IP address against a database of IP-to-company mappings. So, instead of "someone visited your pricing page," you get "someone at a 200-person manufacturer visited your pricing page." It won't tell you which person, but it tells you the company, and that's enough to route the visit to a sales rep or score it against your ICP.

That's a different thing from personal identification, which tracks the individual. Our guide on how to identify anonymous website visitors explains the difference and how each method works.

How Leadfeeder closes the gap

Leadfeeder has two products that read your website at company level.

Web Visitors shows the companies by name. You stop looking at 8,000 anonymous sessions and start seeing that a 200-person manufacturer visited your pricing page twice this week. That's the kind of signal you can hand directly to a sales rep. To see which companies are already visiting, start with website visitor identification and filter by your target market.

Leadfeeder Web Insights adds company-level data to your traffic reports. Instead of "200 sessions from LinkedIn this month," you see how many companies each channel sent and how many of them fit your ICP. The ICP ratio, the share of identified companies that fit your ideal customer profile, is the number that tells you whether a channel is sending companies you sell to. One tracking script, ready in minutes, and the reports are ready out of the box. On paid plans from Discover up, you can type a question in plain English and get a full report back. Install the tracking script and open the Traffic Sources report after a week to see how many of the companies each channel sent fit your ICP.

See which companies are already on your site and what brought them, with a 14-day free trial and no credit card. Terms and conditions apply.