You have taken the job. There are four people in the team, eleven channels that somebody is publishing to, a SharePoint intranet halfway through a redesign, and a Viva Engage estate of ninety-odd communities that nobody has looked at as a whole. In roughly six weeks the CCO will ask what you have found, and in month five there is a budget review. What you audit in the first ninety days determines whether that budget conversation is a defence or a negotiation.
The mistake is to start with strategy. A strategy written before you know how many people in your organisation do not have a work email address is a document about an organisation you have imagined. This is a sequence, and the order matters more than the content of any single step.
Days 1 to 15: inventory the estate, touch nothing
Produce one page listing every surface on which internal communication currently reaches somebody. Not the surfaces in the strategy deck. The ones actually in use. In a Microsoft 365 organisation that usually means the SharePoint intranet and its news sites, Viva Engage communities, Teams channels being used for communication rather than project work, Viva Connections on mobile, the newsletter platform, digital signage in plants, and the manager cascade that nobody owns but everybody relies on.
Against each one, record three things: who owns it, who publishes to it, and when anyone last looked at a number for it. That third column is the finding. In most organisations, between a third and a half of the estate has never been measured by anyone, and several surfaces will turn out to have an owner who left.
Resist the urge to rationalise the estate in week two. You do not yet know which of those unloved channels is the only thing reaching the night shift.
Days 16 to 30: establish the denominator
This is the step almost everyone skips, and skipping it is unrecoverable in the sense that every number you produce for the next two years will be uninterpretable without it.
Go to HR and to whoever administers Entra ID[1] and build the workforce picture: headcount by function, by site, by country, by language, by contract type, and, critically, split by whether the person has a work email address and a device they read it on. That last split is the single most useful number a new Head of Internal Communication can produce in month one. In a manufacturer or a retailer it is frequently between 40 and 70 per cent of the workforce, and it reframes every subsequent conversation about reach.
Gallup’s workplace research has consistently placed the widest engagement gaps in frontline and deskless populations[2], and Deloitte’s human capital work keeps returning to the same disconnect between corporate intent and frontline experience[3]. You are not going to close that in ninety days. You are going to size it, which is a different and achievable thing. Our note on audience segmentation covers how to structure the attributes so they stay usable.
Days 31 to 50: capture the baseline before it disappears
Native reporting retains a finite window. Whatever history exists today in SharePoint site analytics[4], Viva Engage community insights and your newsletter platform will roll off, and once it has, nobody can reconstruct it. Export everything now, even in a form you are not happy with, and write down each source’s retention limit next to it.
Two specific things to capture. First, twelve months of channel-level activity, whatever your tools hold. Second, the three or four campaigns your predecessor considered significant, at post level, because you will want to reference them and the detail will be gone in a quarter.
Do not attempt to fix the numbers at this stage, and do not present them. A baseline that says “here is what the previous system could see” is more valuable than a cleaned-up figure that quietly excludes the channels you have not worked out yet. The orientation in our guide to internal communication analytics is a reasonable companion at this point.
Days 51 to 70: talk to the people the data cannot see
Twelve to fifteen conversations, thirty minutes each, chosen deliberately from the parts of the organisation your channel data is thinnest on. A plant manager. A shift supervisor. A regional HR business partner. A field engineer. The executive assistant who forwards things to a distribution list because the official channel does not reach her director’s team.
Ask one question in different forms: where did you hear about the last significant company announcement? The answers will name channels that are not on your one-page inventory, and that is the point. In one 12,000-person manufacturer we worked with, the most reliable route to the shop floor turned out to be a printed sheet a supervisor pinned to a noticeboard after reading the SharePoint article on his phone in the car park. No dashboard was ever going to surface that.
| Phase | What you produce | Where it comes from | Typical trap |
|---|---|---|---|
| Days 1 to 15 | Channel register with owner and last-measured date | Interviews, tenant admin, publishing calendars | Rationalising before you understand coverage |
| Days 16 to 30 | Workforce denominator by function, site, contract, email access | HR extract and Entra ID | Using channel membership as a population |
| Days 31 to 50 | Twelve-month baseline plus retention limits per source | SharePoint, Viva Engage, newsletter platform | Waiting until the data has rolled off |
| Days 51 to 70 | Qualitative map of unofficial routes | 12 to 15 structured conversations | Interviewing only head office |
| Days 71 to 90 | One baseline, one gap, one commitment with a number | All of the above | Presenting a strategy instead of a finding |
Days 71 to 90: one baseline, one gap, one commitment
Do not present a strategy at day ninety. Present three things.
The baseline: here is what we can currently see, here is what we cannot, and here is the retention limit that means we will lose more of it. The gap: here is the population we are not reaching, sized. The commitment: one measure, with a starting number and a target date, that you are prepared to be held to.
One commitment, not five. A new Head of Internal Communication who promises to lift an engagement score has promised something they do not control. A new Head who says “17 per cent of production employees saw the last three safety communications within seven days, and in twelve months that will be above 50” has promised something specific, measurable and theirs. Our piece on executive reporting for internal communications covers how to frame that conversation, and internal communication KPIs for 2026 sets out the measures worth committing to.
Three traps in the first ninety days
Rebuilding the intranet. It is the most visible thing you could do and it will absorb the entire first year. If a redesign is already in motion, your job in the first ninety days is to make sure it is measured before and after, not to take it over. Our note on measuring after a SharePoint redesign explains what to capture in advance.
Buying tooling in month one. You do not yet know your questions. Native analytics is the correct starting point: SharePoint site analytics and Viva Engage community insights are free, already switched on, and entirely sufficient for auditing single surfaces, which is what the first ninety days consists of. Buy when the questions have outgrown what native can answer, not before.
Promising a number you do not own. Engagement scores move with pay reviews, restructures and line management quality. Commit to reach and coverage in populations you can name. Those are yours.
A note on tooling, once the ninety days are up
The questions that outgrow native are predictable: anything phrased by function or site rather than by channel, anything spanning more than one surface, and anything requiring a trend longer than the retention window. At that point there are two routes.
A custom Power BI build on the Microsoft Graph reporting API is the first one most teams consider. It can work. Go in knowing that the reporting API is not consistently reliable and produces days with missing data, that customers have observed KPI values in Power BI that do not reconcile with the native SharePoint reports for the same period, and that a dependable build needs a robust ingestion process, scheduled reconciliation against the native figures and a named owner. Budget for the pipeline, not the dashboard. The detail is in our honest guide to SharePoint analytics in Power BI.
The other route is a dedicated measurement layer. Tryane reads SharePoint, Viva Engage, Teams and newsletter platforms together, joins that activity to your organisational structure through Entra ID or an HR file, and keeps unlimited history across product transitions, which is exactly the retention problem described in the day 31 to 50 phase. SOC 2 Type 2 certified, GDPR compliant by design, EU hosting by default with US data residency on request, single sign-on through Azure AD or Entra ID, deployed in a couple of hours with no agent to install.
Frequently asked questions
What should a new Head of Internal Communication do in week one?
Build the channel register, not the strategy. List every surface actually reaching people, with its owner and the date anyone last looked at a number for it. It takes about three days of interviews and it is the document everything else in the first ninety days depends on.
Why establish the workforce denominator before measuring anything?
Because reach is a fraction and channel data only gives you the numerator. Without a population defined by function, site, contract type and email access, every reach figure you produce is a count with no meaning, and the populations you are failing to reach stay invisible by construction.
How much history can we recover if we start late?
Less than you would hope. Native reporting keeps a finite window, so history that has rolled off cannot be reconstructed after the fact. This is why the baseline export sits at day 31 rather than day 200, and why teams that inherit a measurement gap usually have to start their trend line from the day they arrived.
Should the ninety-day review include a channel rationalisation plan?
Only if you have coverage data for the channels you propose to close. Closing a low-volume channel that happens to be the only route to a night shift is a costly way to demonstrate decisiveness. Size the audience of each surface first, then rationalise in month six with evidence.
Do we need a dedicated analytics tool to run this audit?
No. The ninety-day audit is deliberately designed to run on native tooling plus an HR extract, because at this stage your questions are single-surface. The case for a dedicated layer appears later, when questions start crossing channels and populations. Our guide to choosing an internal communications analytics tool covers that decision when you reach it.
Sources
• Microsoft Learn, what is Microsoft Entra
• Gallup, State of the Global Workplace
• Deloitte, Human Capital Trends
• Microsoft Learn, SharePoint portal health and site analytics
Further reading
• Building an internal communications measurement strategy
• A guide to internal communication analytics
• Audience segmentation for internal communications
• Measuring frontline worker communications
• Internal communications benchmarks for 2026
Tryane runs a 15-minute working session with Heads of Internal Communication, and it is a useful thing to do in the first ninety days: we walk through your channel register and show which parts of your workforce your current data can and cannot see. Book a slot with Jérémy to schedule yours.