Your tenant has 214 communities. Sixty-one of them had a post in the last quarter. Eleven of them account for most of the activity. On your desk this morning is a request from a transformation programme for a new community, and the person asking is senior enough that saying no will cost you something. You will probably say yes, and in eighteen months it will be community number 215 in the dormant column.
The argument of this article is that a community is not a container, it is a claim on somebody’s weekly time. Every one you create is a standing commitment by a named human to read it, answer in it and care whether it is alive. Treat the estate as a portfolio of those commitments and the numbers stop being embarrassing, because the question changes from how many communities you have to how many you can actually staff.
What a community actually costs
The visible cost of a community is nothing. Creation is a few clicks, storage is included, and no budget line moves. This is exactly why estates sprawl: the cost is real but it lands on a person rather than a budget.
A community that works consumes something like two to four hours a week of a community manager’s attention: reading, seeding questions, chasing an answer for the person whose post has sat unanswered since Tuesday, and knowing enough about the subject to tell a good thread from a dead one. Multiply that by 214 and the arithmetic tells you immediately why most of the estate is dormant. Nobody decided to abandon those communities. There was simply never enough attention to go round, and attention is the input that cannot be provisioned.
This reframing is what makes the rest tractable. You are not deciding whether a topic deserves a community. You are deciding whether it deserves a slice of a finite pool of human attention, in preference to whatever that attention is currently doing. The article on working with community managers covers who those people are and how to support them.
The launch gate: four questions
Before a community is created, four things have to be true. Write them into the request form and the sprawl stops within a quarter.
Is there a named owner who has agreed, in writing, to two hours a week? Not a team, not a mailbox. A person. If the sponsor cannot find one, the answer is no, and this alone will decline roughly half of incoming requests without any argument about the merits of the topic.
Does it need to cross a boundary? If everyone who would participate already sits inside one team, the request is for a Teams channel and it will work better as one. This is the single most common misdirected request, and our comparison of Teams and Viva Engage is the document to hand people when the conversation gets circular.
Can the owner state its purpose in one sentence a stranger could act on? “A place for updates about the programme” fails, because it describes a broadcast. “Where site quality leads ask each other about audit findings before escalating” passes, because it tells a stranger whether they belong.
Is there a first question, ready now? Not an announcement. A question the owner genuinely does not know the answer to. Communities that open with a welcome post and no question start dead and rarely recover.
The first 90 days
By day 90 three things have to be true, and if they are not, the honest move is to close rather than to extend the runway. There must be replies from more than one function or site, because a community answered only by the owner’s own team is a Teams channel with more steps. There must be at least a handful of contributors who are not the owner and not the sponsor. And there must be a question that was answered by someone the asker did not know, which is the specific thing this channel exists to produce and the thing our piece on which use cases survived into Viva Engage examines in detail.
Ninety days is deliberately short. Long runways do not rescue communities, they only make the eventual closure more awkward, because by then the sponsor has mentioned it in a steering pack.
| Stage | The question | The evidence | The decision |
|---|---|---|---|
| Gate | Should this exist at all? | Named owner with committed hours, cross-boundary need, one-sentence purpose, a real first question. | Create, or redirect to a Teams channel. |
| Day 90 | Did it take? | Replies from more than one function or site, contributors beyond owner and sponsor, one question answered by a stranger. | Continue, merge into a broader community, or close. |
| Steady state | Is it broadening or narrowing? | Distinct contributors per quarter, top-ten concentration, answered-question rate. | Continue, or intervene on concentration. |
| Drift check | Is it still what it said it was? | Whether recent posts match the stated purpose. Annual review. | Restate the purpose, split, or merge. |
| Sunset | Has it finished? | Two consecutive quarters below the launch bar, or the programme it served has ended. | Archive with a closing post, keep the content searchable. |
The two ways a healthy community fails
Concentration. Activity holds up while the number of distinct contributors quietly falls, because four enthusiasts are carrying it. The aggregate looks fine, so nobody investigates, and then two of the four change roles and the community drops off a cliff in a single month. Track distinct contributors per quarter, not post volume, and treat a rising top-ten share as the warning it is.
Drift. A community created for site quality leads gradually becomes the place the quality function posts its announcements. Nothing dramatic happens. The purpose statement stops describing the content, the cross-boundary replies stop, and it becomes a newsletter with a comment box. The annual drift check exists to catch this, and the fix is usually to restate the purpose publicly rather than to close.
Closing a community well
Closure is the discipline almost nobody has, and it is the one that determines whether the estate stays credible. Three points make it survivable.
Say it out loud. A closing post that names what the community was for, thanks the people who carried it and says where the conversation moves next costs ten minutes and prevents the impression that things silently rot here.
Archive, do not delete. A dormant community frequently holds a genuinely useful record of how a problem was solved, and destroying it removes institutional memory nobody will miss until they need it. Microsoft 365 group expiry policy will act on inactive groups automatically if you let it[1], which is a reason to make the decision deliberately rather than discovering it has been made for you. The mechanics of what changed in this area after the rename are set out in our article on the difference between Viva Engage and Yammer.
Move the people, not just the content. If a community closes because it is merging into a broader one, post the redirect inside the old community rather than assuming the membership will find its way. Most will not.
Measuring the portfolio rather than the community
Community-level insights answer community-level questions, and Viva Engage provides them natively[2]. If you run one community and want to know whether it is alive, that is free, already switched on, and sufficient. Portfolio questions are different in kind. How many communities cleared the day-90 bar this year? Which functions have no live community at all? Is participation broadening across the estate or concentrating into the same eleven?
Those questions need activity joined to organisational structure, which the native panels do not hold, and they need history that survives the retention window. Gallagher’s State of the Sector research keeps finding measurement to be the capability internal communicators name as their weakest[3], and portfolio reporting is where that bites hardest, because the aggregate view is precisely what native tooling was never built to produce.
A Power BI build against the Microsoft Graph reporting API is the usual attempt at a workaround. It can be made to work. It is also more demanding than it looks: 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, and a dependable build needs a robust ingestion process, scheduled reconciliation against native figures and a named pipeline owner. Budget for the pipeline rather than the dashboard, and read our honest guide to SharePoint analytics in Power BI first.
Tryane provides the portfolio view. It reads Viva Engage alongside SharePoint, Teams and your newsletter platform, joins activity to your organisational structure through Entra ID or an HR file so you can see which functions and sites are actually represented, and keeps history across product transitions so a community that started life in Yammer still trends as one line. It is SOC 2 Type 2 certified and GDPR compliant by design, hosts in the EU by default with US data residency on request, deploys in a couple of hours over Azure AD or Entra ID single sign-on, and installs no agent. For what a single community manager should have in front of them, see what a Viva Engage community dashboard should contain.
Frequently asked questions
How many communities should a 10,000-person organisation have?
There is no correct number, and any benchmark you are quoted is someone else’s staffing model. The useful constraint is how many community managers you have and how many hours each can genuinely give. If that pool supports twenty properly run communities, twenty is your number, whatever the current count says.
What if a senior sponsor insists on a community that fails the gate?
Offer the Teams channel and be specific about why. Most of these requests are really requests for a place to coordinate, and the sponsor does not care which product provides it. The ones that survive that conversation usually turn out to have a genuine cross-boundary need, which means the gate did its job.
Should dormant communities be deleted?
Archive rather than delete, unless there is a records-management reason to remove the content. Dormant communities often contain answers people still search for, and once deleted that knowledge is gone. Closing publicly and leaving the archive searchable gives you the tidiness without the loss.
Is low activity always a problem?
No. A community of forty specialists that produces one genuinely useful exchange a month can be more valuable than one of four thousand producing daily noise. Judge against the stated purpose, not against a volume target. Gallup’s research on engagement and performance points at the quality of specific work relationships rather than at communication volume[4].
Does Tryane replace Viva Engage analytics?
No. Viva Engage keeps recording what it records. Tryane sits above it and provides the estate-level view, joining activity across communities and channels to your organisational structure so portfolio questions become answerable. You keep the Microsoft 365 investment and add the layer native reporting does not provide.
Sources
• Microsoft Learn, Microsoft 365 group expiration policy
• Microsoft Learn, Viva Engage analytics
• Gallagher, State of the Sector
• Gallup, engagement and business outcomes
Further reading
• Viva Engage in 2026, what it is actually for
• The first 90 days for a new Viva Engage owner
• What to post in Viva Engage, and how often
• Creating communities people actually join
• What Yammer changed, and what Viva Engage broke in the measurement
Tryane runs a 15-minute working session with Heads of Internal Communication, looking at which of your communities cleared the day-90 bar and which are being carried by four people. Book a slot with Jérémy.