Open the migration plan for a Workplace to Viva Engage move and you will find rows for community mapping, content export, account provisioning, governance model, training and a launch communication. What you will not find is a row that says “prove we did not lose the 4,000 people who only ever used Workplace on their own phone”. That row is the one the CCO will eventually ask about, usually about eight months after go-live, usually in the same meeting as the budget.
Meta wound Workplace down in phases: normal service ended, a read-only period followed, and after that the data was scheduled for deletion. Most organisations on it moved to Viva Engage, which is the sensible destination for a Microsoft 365 estate. This article is not about whether that was the right call. It is about the measurement work that sits either side of the move, and the one structural risk that content migration plans routinely miss.
The identity boundary is the real risk
Workplace and Viva Engage differ in a way that matters more than any feature comparison. Workplace was a standalone product with its own account model, and in a lot of organisations that is exactly why it reached the frontline: a warehouse operative or a store colleague could get an account and use it on a personal device without a corporate mailbox sitting behind it.
Viva Engage is part of Microsoft 365[1], and access rides on tenant identity through Entra ID[2]. That is a genuine advantage for governance, security and integration with the rest of the estate. It also means that “did we lose the frontline” is really three questions in a trench coat: did those people get an identity in the tenant, do they have a route to sign in on the device they actually own, and did anyone check.
The pattern we see is not a dramatic failure. It is a quiet one. Head office moves across in a fortnight, the numbers look healthy because head office generates most of the visible activity anyway, and the population that Workplace was quietly serving simply stops appearing. Gallup’s workplace research has consistently placed the widest engagement gaps in frontline and deskless populations[3], which is precisely the group an aggregate activity number will not surface.
What to preserve, and what it is actually for
Three assets, in descending order of how often they are forgotten.
The membership map. Workplace group membership was a segmentation of your organisation that people built by hand, by choosing what they cared about. It is a better audience map than anything in your channel strategy document, and it is the thing most migrations discard because the project scope says “recreate the groups”, not “keep the membership data”. Export it as a list, not just as a set of recreated communities.
The measurement baseline. Twelve months of activity by group, and if you can get it, by population. This is the only version of “before” that will ever exist. Once the source system is gone, no reconstruction is possible from either side.
The archive of problem solving. The threads where a technician in one region answered a question from a technician in another. Most organisations export content wholesale and never look at it again, which is fine, but be deliberate about which threads are worth surfacing in the new environment rather than dumping into cold storage.
| What to preserve | Why it matters | Where it goes | If you skip it |
|---|---|---|---|
| Group membership lists | A hand-built audience segmentation | Audience model, joined to HR attributes | You rebuild communities blind |
| 12 months of activity data | The only “before” that will exist | Measurement layer or an offline archive | No before-and-after is possible |
| Top-performing content | Evidence of what the audience responds to | Editorial planning for the new estate | You restart the learning curve |
| Frontline access patterns | Device and identity routes that worked | Provisioning and enablement plan | The identity boundary bites silently |
| Named community managers | The people who generated most activity | Owner mapping in Viva Engage | Activity drops and nobody knows why |
The three measurements that answer the question
Population coverage, same denominator both sides
Not total active users. The share of a named population that was active in a comparable period, before and after. The denominator has to be identical on both sides and it has to come from HR or Entra ID rather than from either platform’s membership. “Manufacturing headcount in France” is a denominator. “People in the community” is not.
Participation rate by segment, never the aggregate
Run the same measure by function, site, country and contract type. A migration that holds total activity flat while shifting it from 40 per cent of the field force to 4 per cent has failed, and the aggregate will show a flat line the whole way. Structuring these attributes properly is covered in our note on audience segmentation for internal communications.
Time to first action after provisioning
For migrated users, how long between the account existing and the person doing anything at all. This is the metric that separates a provisioning success from an adoption success, and the two get conflated constantly in migration reporting. An account created is not a person reached. If a cohort’s median time to first action is “never”, you have found your gap, and you have found it early enough to fix with an enablement push rather than a post-mortem.
Running the before-and-after honestly
Three rules, and a concession.
Compare the same weeks. Activity in an industrial or retail workforce is shaped by shutdowns, peak trading and school holidays far more than by your migration. Compare the same denominator. And fix the definition of “active” in writing before you look at either number, because the temptation to adjust it afterwards is considerable and everyone in the room will know.
The concession: expect a real drop in the first ninety days regardless of how well you ran the project. People had habits, and the habits were attached to an icon on a phone. A 20 to 30 per cent fall in participation immediately after a platform change is normal and does not mean the migration failed. Set the honest comparison point at six months, tell your leadership that in advance, and you will save yourself an unnecessary crisis in month two.
What is not normal, and what the six-month measure exists to catch, is a drop that is concentrated rather than uniform. If participation fell 25 per cent everywhere, you have a habit problem and a re-engagement campaign. If it fell 5 per cent in head office and 60 per cent in stores, you have an access problem, and no amount of content will fix it. The practicalities of measuring that population are in measuring frontline worker communications.
Where native analytics gets you, and where it stops
Viva Engage community insights will tell you, honestly and for free, how each community is doing after the move. For a community manager watching their own community find its feet, that is sufficient and there is no reason to add anything.
It cannot do the one thing this article is about. A before-and-after across a platform migration requires data from two systems, one of which no longer exists, expressed in one consistent definition, against a denominator that comes from neither. No native panel was ever going to produce that, and it is not a criticism to say so. The same constraint applies to organisations still carrying the Yammer to Viva Engage transition, which we covered in how Yammer influenced your organisation. If you are weighing native against a dedicated layer for this kind of question, the side-by-side in Tryane compared with Viva Engage native analytics sets out which questions each answers.
Some teams attempt the before-and-after in Power BI. It can be built. The Microsoft Graph 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 owner. Budget for the pipeline rather than the dashboard, and note that none of it solves the harder problem, which is that the Workplace side of the comparison has to have been captured before the source went away. Our honest guide to SharePoint analytics in Power BI covers the failure modes.
This is the specific gap Tryane was built for: reading Viva Engage alongside SharePoint, Teams and newsletter platforms, joining that activity to your organisational structure through Entra ID or an HR file, and keeping unlimited history across product transitions so a migration reads as one continuous trend rather than two disconnected reports. 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 happened to Workplace from Meta?
Meta wound the product down in phases: an end to normal service, then a read-only period during which organisations could retrieve their data, then scheduled deletion. Most Microsoft 365 organisations moved to Viva Engage. Our overview of Workplace alternatives covers the wider options.
Can we still recover our Workplace analytics?
Not after the retrieval window closed. This is why the measurement baseline belongs in the migration plan rather than in a follow-up project. If you did not capture it, the practical fallback is to set your new baseline at go-live, state clearly that the before side is unavailable, and measure forward from there. It is a weaker position than a genuine comparison, and it is far better than a reconstructed figure nobody can defend.
Does everyone need a Microsoft 365 account to use Viva Engage?
Access is tied to identity in the tenant, so people who participated on Workplace without a corporate account will need provisioning and a workable sign-in route on the device they actually use. Treat that as an audience question rather than an IT ticket: work out which populations are affected, how large they are, and what device they hold, before go-live rather than after.
How long should we wait before judging the migration?
Six months, with a checkpoint at ninety days that you have already told leadership will look bad. Habit change takes longer than a launch campaign, and the ninety-day number reliably understates where the platform will settle. What the ninety-day checkpoint is genuinely useful for is spotting a concentrated drop in one population early enough to act.
What is the single most useful thing to measure after the move?
Participation rate by population against a denominator from HR or Entra ID, compared with the same weeks before the migration. One number, segmented. It answers the question leadership is actually asking, and it is the number that reveals a lost frontline while there is still time to do something about it. The wider set is in internal communication KPIs for 2026.
Sources
• Microsoft Learn, Viva Engage overview
• Microsoft Learn, what is Microsoft Entra
• Gallup, State of the Global Workplace
• Gallagher, State of the Sector
Further reading
• Workplace from Meta alternatives
• Viva Engage in 2026, a complete guide
• Measuring frontline worker communications
• Tryane compared with Viva Engage native analytics
• Audience segmentation for internal communications
Tryane runs a 15-minute working session with Heads of Internal Communication who are mid-migration or just out of one, rebuilding the before-and-after on your own populations so you can see whether the drop is uniform or concentrated. Book a slot with Jérémy to schedule yours.