Intranet analytics: what to measure, which tools, and what is allowed?

An intranet is one of the largest internal communication investments most organisations make. Yet the question “is anyone actually reading this?” is rarely answered with data. Editors steer on gut feeling, communications advisers on reactions at the coffee machine.

That is remarkable, because for the public website measurement has long been standard practice. This guide covers intranet analytics from start to finish: what you can measure, with which tools, and which privacy rules apply. For public sector organisations that last question is not a footnote but the starting point.

Why intranet analytics is a different discipline from web analytics

At first glance it looks the same: page views, visitors, search behaviour. But there is a fundamental difference. The visitors on your intranet are not anonymous website visitors, they are your own employees. They are logged in, identifiable, and employed by the organisation doing the measuring.

That difference has two consequences. Legally: measuring employee behaviour quickly triggers works council consent requirements and needs its own legal basis under the GDPR. Practically: the questions are different. A marketer wants to optimise conversion; an intranet editor wants to know whether the announcement about the reorganisation reached the right departments.

Approach intranet analytics as a copy of web analytics and you run aground on both points. The measurement questions below are therefore ordered around what an editorial team actually needs.

The four measurement areas of intranet analytics

1. Content usage and popularity

The basic question of every editorial team: what is being read, and by whom? Concretely this covers page views and unique visitors per page, the most popular content per team or department, reading time per article, and download statistics for documents.

That last one is often forgotten. On many intranets the real value sits in documents: policies, forms, handbooks. If the staff handbook is never downloaded while HR answers the same question daily, you know where the problem is.

The right comparison is not “page A versus page B” but “reach versus audience”. An all-staff announcement read by 8 percent is a problem; a policy for a team of twenty with fifteen readers is a success.

2. Navigation behaviour and user flow

What do employees click on from the homepage? Which route do they take to frequently used pages? And where do they drop off?

This measurement area exposes structural problems. If the most visited page is only reachable through three intermediate steps, that is a navigation problem no amount of content improvement will fix. The same goes for forms: if 60 percent of users abandon at the same step, that is where the friction sits.

3. Adoption of new functionality

Every intranet renewal introduces new components: a people directory, a knowledge base, an incident form. The question that is rarely asked afterwards: is anyone actually using it?

Measure interactions with new components in the first weeks after launch. Low adoption is not a failure but a signal: the component is hard to find, unclear, or solves a problem nobody had. All three are fixable, but only if you know.

4. Optimising search behaviour

The internal search engine is the most honest feedback source you have. The top ten search terms show what employees are looking for. More important still are the searches with no results: there, people are looking for something that does not exist, or that is named differently than they expect.

Reviewing the failed searches monthly and adjusting content or synonyms accordingly improves an intranet faster than any editorial plan.

Which tools are available?

The tool landscape for intranet analytics is smaller than for web analytics, and the differences sit mainly in privacy and scope.

Built-in statistics (SharePoint, Viva). Microsoft SharePoint has its own usage reports per site and per page. They are usable for a first impression, but the limitations are substantial: history goes back at most 90 days by default, there is no organisation-wide view for editors, and the organisation-wide reports require an administrator role. For the details, see the separate analysis of the built-in SharePoint statistics.

Specialist intranet analytics suites (such as CardioLog Analytics). Extensive packages with heatmaps, engagement campaigns and individual user profiles. Functionally rich, but precisely that individual profiling clashes with the GDPR principles of many European organisations, and the vendors are predominantly American. CardioLog, for example, is a product of the US-based Intlock.

Privacy-first trackers (such as the govanalytics SharePoint Tracker). Measure intranet usage in aggregate, without linking to individual employees. Contextual dimensions such as department or location remain possible, as long as they are not traceable to a person. For public sector organisations this is usually the route that keeps both the editorial team and the data protection officer satisfied.

General web analytics platforms (Matomo, Piwik PRO). Can also measure intranets, provided they are configured correctly for an internal environment. The reporting is more generic, but for organisations already running such a platform it is a logical extension.

What is allowed? The privacy rules in brief

For public sector organisations, and in practice for almost every European employer, three frameworks apply.

Works council consent. Systems capable of monitoring the presence, behaviour or performance of employees fall under works council consent requirements (in the Netherlands: article 27 of the Works Councils Act). Intranet analytics at individual level falls under this, even when monitoring is not the intention.

A DPIA when personal data is involved. Processing employee behavioural data for a new purpose requires mapping the risks in advance with a data protection impact assessment.

Data minimisation. The core question: do you need individual data for your purpose? For editorial decisions the answer is almost always no. Which pages are being read is an aggregated question; who read them does not matter. Measuring in aggregate keeps you outside the heaviest obligations and keeps the relationship with the works council clean.

The practical conclusion: choose a measurement setup that aggregates by default. Then intranet analytics is not a privacy risk but simply an editorial instrument.

Where do you start?

Not with the tool, but with three questions. What do we want to know (pick two or three measurement questions from the areas above, not all four at once)? Who will look at it (a report without an owner goes unread)? And what is allowed within our privacy framework?

With those answers the tool choice is usually quick. And start small: one dashboard with the five most popular pages, the failed searches and the reach of the last all-staff announcement delivers more than forty reports nobody opens.

Frequently asked questions

What is intranet analytics?

Intranet analytics is measuring and analysing how an intranet is used: which pages are read, how employees navigate and search, and which content reaches its audience. The goal is to base editorial and organisational decisions on actual usage instead of assumptions.

Is measuring employee intranet usage allowed under the GDPR?

Yes, provided the setup is right. Aggregated measurement, without traceability to individual employees, is defensible and does not require individual behavioural profiles. Measuring at individual level requires its own legal basis, a DPIA and, in the Netherlands, works council consent.

Does the works council need to be involved in intranet analytics?

For measurement at individual level, yes: systems capable of monitoring employee behaviour require works council consent (article 27 of the Dutch Works Councils Act). For strictly aggregated measurement, where no individual employee is traceable, the situation differs. Aligning with the works council early is sensible either way.

What can you measure on an intranet?

Four areas: content usage (page views, reading time, downloads), navigation behaviour (click paths, drop-off points, form abandonment), adoption of new functionality, and search behaviour (popular search terms and searches without results).

Are the built-in SharePoint statistics sufficient?

For a first impression, yes; for structural editorial work, usually not. History is limited to 90 days, there is no organisation-wide view for editors, and organisation-wide reports require an administrator role in the Microsoft 365 admin centre. Use at individual level also raises privacy questions.