Your Intranet Cannot Fix Engagement. Here Is What It Actually Moves

By Nicolas Saliba · Published December 2025, Updated August 2026 · 8 min read

The annual engagement survey lands four points down, and by the following week someone has asked what internal communications is going to do about the intranet. It is a fair instinct and a bad question. The same quarter carried a reorganisation, a pay round that disappointed two functions, and a change of leadership in the largest business unit. Somewhere in that list is the reason the score moved, and the intranet is not obviously on it. Accepting that openly is not a retreat. It is the only position from which you can make a claim about communication that survives scrutiny, and it is where this article starts.

Engagement has many owners, and you own two of the inputs

Gallup has spent decades establishing what actually predicts engagement, and the picture is consistent: the immediate manager, clarity about what is expected, recognition, development, and whether the work connects to something the person believes matters^[1]. Communication appears in that picture, but as one contributor among several rather than the lever. Their global workplace research has also found for years that most of the world’s workforce is not engaged, and that the gap is widest in frontline and deskless populations^[2]. No intranet in the world closes a gap of that shape on its own.

Two things follow, and they point in different directions. The first is that an internal communications leader who promises to move the engagement score has accepted accountability for an outcome largely produced by other people’s decisions. When the score moves the wrong way, that promise is what gets quoted back. The second is that the same leader owns real inputs, and can be precise about them. That precision is worth more in a budget conversation than a bigger claim would be, because it is a claim you can produce evidence for every month rather than once a year.

The four things a digital workplace genuinely moves

Strip out everything the intranet cannot touch and four mechanisms are left. They are worth naming individually, because each one has a different instrument.

Knowing what is going on. Whether a person can say what has changed in the organisation this month without asking their manager. This is the most direct intranet contribution and the easiest to under-measure, because the aggregate always looks healthy while a specific population sits at zero.

Finding what you need to do the job. The policy, the form, the contact, the procedure. Friction here shows up in engagement surveys as frustration with the organisation rather than frustration with a search box, which is why it is chronically misattributed.

Understanding what leadership is trying to do. Line of sight from the strategy to the work. A digital workplace does not create that understanding, but it is where the explanation either reaches people or does not.

Connection beyond your own team. Knowing that a person in another site solved the problem you are looking at. This is the one that decays quietly, and the one Viva Engage communities exist to carry.

Notice what is not on this list. Pay, workload, manager quality, career path and job security are not on it, and no amount of intranet work moves them. Naming that boundary out loud, in the report, is what makes the four claims above credible.

Instrumenting the part that is yours

Each of the four has a signal you can already get from a Microsoft 365 estate, and each one is only useful when read by population rather than in total.

Mechanism	The signal to measure	Where it comes from	What good looks like
Knowing what is going on	Reach of must-know news into each population, not total views	SharePoint news and newsletter activity joined to Entra ID or an HR file	No population below a floor you have agreed in advance
Finding what you need	Searches that return no click, and the same query repeated by the same person	SharePoint search usage reports^[3]	Your top 20 queries all resolve on the first result
Line of sight	Reach and reply volume on leadership content outside head office	Viva Engage and SharePoint, segmented by site and function	Leadership content performs outside HQ, not only inside it
Connection beyond your team	Share of community participation that crosses site or function	Viva Engage community activity joined to organisational attributes	Cross-boundary participation is growing, not just volume

The findability row is the one most teams skip and the one that pays back fastest. Search usage reports will tell you which queries return nothing useful, and the top twenty are almost always the same twenty: expenses, holiday policy, IT support, the org chart. Fixing those is a fortnight of work and it removes a genuine daily irritation. It will not show up in the engagement score as a line item, and it is still worth doing.

What to claim, and what to stop claiming

Take the engagement score off your slide as a key performance indicator and put it back as context. Then write the sentence that replaces it, and write it before anyone asks: communication is one input to engagement among several, and what we can demonstrate is that must-know information now reaches 81 per cent of manufacturing against 47 per cent a year ago, and that the top twenty searches now resolve.

That sentence does something the bigger claim cannot. It is falsifiable, it is monthly, and it is entirely within your control. Gallagher’s State of the Sector research keeps finding measurement at the top of the capability gaps internal communicators name^[4], and a large part of that gap is teams reaching for an outcome measure that was never theirs. The mechanics of reading survey data honestly alongside communication data are covered in our pieces on mistakes in employee engagement analysis and on using both engagement surveys and communications data.

When the intranet genuinely is the problem

The argument runs the other way too, and the honest version has to include it. There are organisations where the digital workplace is a real cause of disengagement, and three signs give it away. The first is a population with no practical access: if a third of your people work a shift without a company device and your intranet assumes a desktop browser, the problem is structural, and Viva Connections in the Teams mobile app exists precisely for that case^[5]. The second is a search function that fails on the queries people actually make. The third is an intranet nobody has owned for two years, where the news is stale and half the links are dead, which teaches people that the channel is not worth checking. Where those conditions hold, fixing the intranet is engagement work, and you should say so plainly. Our piece on measuring frontline worker communications covers the access question in more depth.

The honest position on native analytics

SharePoint site analytics and the Viva Connections usage view are free, already switched on, and enough if you own one site and want to know whether your pages are read. For a single site owner that is the right tool and there is nothing to add.

Everything in the table above breaks that boundary, because every row asks a question about a population rather than a page. Native reporting does not join to workforce attributes, so it cannot tell you that manufacturing sits at 47 per cent while the commercial function sits at 90. It retains a finite window, which is awkward when the argument you are making is about a year of improvement. And it reports each product separately, while people receive your message across several. Tryane exists in that gap: it reads SharePoint, Viva Engage, Teams and your newsletter platform together, joins the activity to organisational structure through Entra ID or an HR file, and keeps history across product transitions. It is SOC 2 Type 2 certified and GDPR compliant by design, hosts in the EU by default with US data residency on request, and deploys in a couple of hours over single sign-on with no agent to install. For structuring the attributes first, see our note on audience segmentation.

Frequently asked questions

Does a better intranet improve employee engagement?

It improves specific things that contribute to engagement: how well informed people are, how easily they find what they need, how visible leadership intent is, and how connected they feel to colleagues elsewhere. Whether the engagement score moves depends on far more than that, particularly the manager relationship and the terms of the job. Claim the mechanisms, not the score.

Our engagement score fell. Can we tell whether communication was a factor?

Not from the score alone, and not at organisational level. What you can do is compare the movement across populations against communication reach in those same populations. If the sites where reach collapsed are also the sites where the score fell hardest, that is worth investigating and worth stating as a correlation. If they are unrelated, you have learned something more useful: look elsewhere.

Which intranet metrics actually matter for this?

Four, and none of them is total page views. Reach of must-know content by population, search resolution on your top queries, reach of leadership content outside head office, and the share of community participation that crosses a site or function boundary. Our guide to intranet KPIs covers how to present them.

How do we measure people who never open the intranet?

Start by counting them, which most organisations have never done. Join your identity directory to the intranet activity and the population with zero activity in 90 days is immediately visible, usually concentrated in specific sites and contract types. Then decide whether that is an access problem, a relevance problem or a device problem, because the fix differs in each case.

Does Tryane replace our engagement survey?

No, and it should not. A survey measures how people feel, which no activity data can tell you. Tryane measures what reached whom, which no survey can tell you accurately. They answer different questions on different cadences, and the useful work happens when you read them side by side at segment level rather than as two organisation-wide numbers.