You have probably had the meeting where someone senior looks at the monthly communications deck and asks, politely, what they are supposed to do with it. It is a fair question. The deck has eleven charts. Two of them are things the executive committee genuinely cares about, five are things the communications team needs on a Monday morning to run the week, and four are there because someone asked for them in 2023 and nobody has dared remove them since.
The usual reaction is to argue about which KPIs are the right ones. That argument never resolves, because it is the wrong argument. There is no single set of internal communication KPIs that matters. There are at least three sets, each belonging to a different audience with a different question and a different tolerance for nuance, and the reason reporting fails is that organisations try to serve all three from one document.
Three audiences, three genuinely different questions
Start by naming who consumes internal communications data in your organisation. In a 1,000 to 50,000-person business, it is almost always these three, and they are not variants of each other.
The executive committee and the CFO
Their question is not “how is internal comms doing”. It is “is this function reducing a cost or a risk that appears on my agenda”. They read quarterly at best, they will hold three numbers in their head, and they are highly sensitive to a number that moves for reasons nobody can explain. Deloitte’s human capital work repeatedly finds that workforce initiatives lose executive sponsorship when the reporting cannot be tied to an operational outcome[1]. What lands with this audience is coverage of the populations that matter to a specific business event, and the direction of travel over four quarters. What does not land is a channel engagement rate.
The campaign sponsor
The HR director running benefits enrolment, the safety director issuing a mandatory update, the transformation lead six months into an ERP rollout. Their question is narrow and time-boxed: did my audience receive this, and did they act. They want reach inside their population specifically, how long it took to get there, and where the tail is. They do not want your intranet’s overall monthly active users, and showing it to them wastes the only meeting you get.
The internal communications team
Your own question is different again: what should we do differently next week. That needs channel-level diagnostics at a weekly cadence. Which publishing slot performed, which sites are dead, which headline formats get a second click, whether the Viva Engage community that was thriving in March has gone quiet. This set is genuinely operational, it is allowed to be noisy, and it should almost never leave the team.
What belongs in each set
The table below is the split we see working. The right-hand column matters as much as the others: most reporting problems are caused by including things, not by omitting them.
| Audience | The question they are really asking | Measures that answer it | Cadence | Leave out |
|---|---|---|---|---|
| Executive committee, CFO | Is this reducing a cost or a risk I own? | Coverage of critical populations; time to reach 80 per cent on a mandatory message; trend over four quarters; one number tied to a named business event | Quarterly | Channel engagement rates, post counts, sentiment scores without a baseline |
| Campaign sponsor | Did my people get it, and did they act? | Reach within their population; time to reach; action completion where an action exists; the size and shape of the tail | Per campaign, plus one follow-up | Tenant-wide totals, anything about other campaigns |
| Line managers as cascade points | Do I need to say something to my team? | Their team’s read rate versus the site average; the two or three items still unread | Weekly digest | Individual-level data, comparative league tables between managers |
| The IC team | What do we change next week? | Performance by publishing slot, channel and format; dead site list; community activity trend; search queries with no result | Weekly | Nothing. This set is allowed to be messy |
Notice that the same underlying event feeds several rows. “2,400 people opened the safety bulletin” becomes, for the safety director, “94 per cent of plant operators within 48 hours, with the Poland sites at 61 per cent”. For the executive committee it becomes a line in a four-quarter trend on mandatory message coverage. For your own team it becomes “the Tuesday 07:00 slot beats the Thursday 14:00 slot on operations”. One event, three framings. The measurement layer has to hold the event once and slice it three ways, which is the practical reason the split is hard without segmentation.
Why the single-deck habit persists
Because it is the path of least resistance. One deck, produced once, circulated widely, feels efficient. It is not, and the failure mode is predictable: the deck grows every quarter. Nobody’s question is fully answered, so a chart gets added rather than swapped. Within two years the pack has twenty charts and no argument, and the executive committee has quietly stopped reading it.
Gallagher’s State of the Sector research has for years found measurement to be the capability internal communicators name as their weakest[2]. Read carefully, the gap is often not analytical skill. It is editorial discipline about who each number is for. If you are staring at a bloated scorecard right now, our companion piece on how to decide which internal communication KPIs to drop is the practical pruning method, and the 2026 KPI pillar sets out the five indicators worth keeping when you rebuild.
The one number that has to be shared
There is a caveat to all this splitting. One measure should appear in every set, defined identically: reach within the addressable population, expressed as a percentage rather than a count. It is the common currency. When the campaign sponsor and the executive committee are both looking at a percentage of a defined population, an argument about a number becomes an argument about the definition of the population, which is a far more productive argument to have.
The honest position on native analytics
The Microsoft 365 admin centre activity reports are free, switched on, and adequate for the third audience on the list, your own team[3]. If you want to know which SharePoint site was busy last week or how many people posted in a Viva Engage community, native reporting answers that well and you do not need anything else.
The split described above breaks native reporting for the first two audiences, and for one specific reason: both need a denominator. “94 per cent of plant operators” requires knowing who the plant operators are, which means joining channel activity to function, site, country and contract type. Native reports do not hold workforce attributes, and their retained window makes a four-quarter executive trend awkward. That is a structural limit, not a criticism.
Many teams attempt the join in Power BI against the Microsoft Graph reporting API[4]. It is a legitimate route with a real cost. The reporting API is not consistently reliable and produces days with missing data. Customers have observed KPI values in Power BI that do not reconcile with the native SharePoint reports for the same period, which is exactly the failure you cannot afford in front of the audience that reads quarterly. A dependable build needs a robust ingestion process and scheduled reconciliation against native figures, with a named owner. Budget for the pipeline rather than the dashboard. Our honest guide to building SharePoint analytics in Power BI covers the specific failure modes.
Where Tryane sits
Tryane reads SharePoint, Viva Engage, Teams and your newsletter platform together, joins the activity to your organisational structure through Entra ID or an HR file, and holds history across product transitions. That is what makes one event serviceable to three audiences: the same read is expressible as a percentage of plant operators, as a line in a four-quarter trend, and as a publishing-slot diagnostic. For the executive-facing half of the split, the board pack that earns budget shows the format, and how to calculate the ROI of internal communications covers the arithmetic behind the one number tied to a business event.
Tryane 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 with single sign-on through Azure AD or Entra ID.
Frequently asked questions
How many KPIs should an internal communications team report?
Per audience, not in total. Three to five for the executive committee, three or four for a campaign sponsor, and as many as are useful for your own team. Counting the total across all three is what produces the twenty-chart pack. If a measure does not have a named audience and a stated decision it informs, it does not belong on a report at all.
Which internal communication KPI does a CFO actually respond to?
Coverage of a population tied to a business event, expressed as a percentage with a trend. “Ninety-one per cent of operations staff reached within 48 hours on mandatory safety messages, up from 74 per cent a year ago” is a sentence a finance audience can use. An engagement rate on its own is not, because there is no baseline to judge it against and no cost or risk attached to it.
Should line managers see communications data about their teams?
A limited version, yes, and it changes behaviour more than most executive reporting does. Give them their team’s read rate against the site average and the two or three items still unread. Do not give them individual-level data or a league table against other managers. Both turn a useful nudge into a surveillance problem and will end the programme quickly.
How often should each set be reported?
Weekly for the team, per campaign plus one follow-up for sponsors, quarterly for the executive committee. Reporting to the executive committee monthly is a common mistake: it produces noise, invites questions about variance that has no cause, and burns the attention you need when something genuinely moves.
Can I produce all three sets from Microsoft 365 native reporting?
The operational set, yes, comfortably. The other two need a denominator drawn from workforce attributes, and native reporting does not hold those. You will either export and join the data yourself on a recurring basis, or use a measurement layer that performs the join and keeps the history.
Sources
• Deloitte, Global Human Capital Trends
• Gallagher, State of the Sector
• Microsoft Learn, Microsoft 365 activity reports
• Microsoft Learn, Microsoft Graph reporting API
Further reading
• Internal communication KPIs for 2026: the five indicators that show your IC is working
• Internal communication KPIs: how to decide which ones to drop
• Executive reporting: the board pack that earns budget
• Audience segmentation for internal communications
• An internal communications reporting template you can reuse every month
Tryane runs a 15-minute call with Heads of Internal Communication, splitting your current reporting pack into the three audience sets and showing which numbers you cannot yet produce. Book 15 minutes with Jérémy.