Microsoft released a free tier of Teams last month, and the interesting thing about it is not the price. It is what the Teams bundling strategy does to every collaboration renewal conversation from here. Bundling has never been an argument about features. It is an argument about arithmetic, and the arithmetic is now uncomfortable enough that finance departments who have never opened a messaging product are going to start asking what the line item is for. Slack pricing pressure does not arrive as a competitor winning a bake-off. It arrives as a question at renewal that nobody has a clean answer to: what are we paying for, given what we already own?
There are two prices in every collaboration deal
The first is list price. The second is marginal price, and only the second one matters. An organisation on Office 365 E3 has already paid for Teams. The seat is bought, the tenant exists, the identity is provisioned, and the admin console is already someone's job. The marginal cost of switching messaging on is close to zero, and the licence is sunk into a subscription that will be renewed regardless of what happens to the chat tool. A paid Slack workspace sitting on top of that carries its full price. Standard runs around $6.67 per user per month billed annually, Plus around $12.50. For five hundred people that is somewhere between $40,000 and $75,000 a year, on top of a bill that is already paid, for a category the organisation already has a product in. That comparison has existed since Teams launched. What changed last month is that the same arithmetic now extends downward to organisations that never bought Office 365 at all. The free tier covers up to 300 users with unlimited chat messages and unlimited search, ten gigabytes of shared file storage plus a couple of gigabytes per person, guest access, audio and video calling, screen sharing, and the browser versions of Word, Excel and PowerPoint. Read that list against how free messaging tiers normally work. The standard design gives away the product and withholds the thing that makes the product valuable over time, which is history. A ten-thousand-message searchable limit across an entire workspace is not a storage constraint, it is a conversion mechanism: it is reached in weeks by any team doing real work, and the upgrade is bought by people who have just discovered they cannot find last quarter's decision. Microsoft's free tier removes precisely that constraint. It is a targeted move, and the target is the conversion path rather than the installed base.
What bundling actually attacks
Not the product. The budget line. The first effect is a shift in the burden of proof. Before bundling, the incumbent messaging tool was renewed because it worked and people used it. After bundling, it has to be justified against something the organisation is already paying for and is not going to stop paying for. "It is better" is a real answer, but it is a harder answer than "it works", and it has to be given to someone who uses neither product. The second effect is a ceiling on price. A standalone product competing against a bundled one cannot charge much more than the difference in value it delivers over the bundled alternative. That difference is real today, and it is a smaller number than the gap between zero and full list price. Price ceilings do not show up as lost deals. They show up as deeper discounts, shorter terms, and accounts that renew flat while headcount grows. The third effect is a change in who decides. A per-team purchase on a corporate card is a preference decision. A five-figure annual line item on a consolidation slide is a procurement decision, and procurement is structurally biased toward fewer vendors, fewer contracts and fewer renewal dates. This is the part that catches product teams out: the tool does not have to lose on merit, it only has to appear on the wrong slide.
The saving that does not show up in the model
Consolidation business cases are reliably wrong in the same direction, because the saving is easy to calculate and the cost is not. The licence saving is a single number, visible, and defensible in a meeting. The migration cost is distributed across exporting or abandoning message history, rebuilding integrations that quietly carry operational work, retraining people who were fluent and are now not, running two systems in parallel for a quarter because nobody trusts a hard cutover, and the administrative time to rebuild channel structure, permissions and guest access. None of that appears on the slide, and most of it lands on teams who were not in the room. The second modelling error is treating the bundle as complete. It is not. Dial-in audio conferencing, the phone system capability and calling plans are separate SKUs. The advanced compliance and eDiscovery features a regulated buyer assumes are included sit in the higher tier. Storage beyond the included allowance is chargeable. A case built on "we already own it" frequently discovers halfway through that the version it already owns is not the version the business case assumed.
| Cost category | What the review must include |
|---|---|
| Marginal licence | What is already contracted and what is actually additional |
| Content and integrations | History treatment and the workflows that must be rebuilt |
| People and cutover | Retraining, administration and parallel running |
| Bundle gaps | Required conferencing, telephony, compliance and storage entitlements |
| Group commitments | Renewal dates, entity agreements and workforce changes |
Qualitative summary of this article's source text, not a measured outcome or performance estimate.
Practical Guidance for Collaboration Licensing Review
- Establish marginal cost, not list price. Work out what each option costs given what is already contracted and cannot be cancelled. Comparing two list prices produces a conclusion that has nothing to do with your cash position.
- Price the migration before you price the licence. History, integrations, retraining, parallel running, admin time. If migration exceeds two years of licence saving, the saving is not a saving, it is a deferral with extra steps.
- Check precisely what the bundle does not include. Conferencing, telephony, the compliance tier, storage above the allowance. Map current usage against the SKU you actually own, not the SKU in the marketing comparison.
- Count the integrations that carry work, not the ones that exist. Most workspaces have dozens of connected apps and three or four that operational processes depend on. Rebuilding three is a task. Discovering the fourth in production is an incident.
- Model the licence at three headcounts. Today, plus twenty per cent, minus twenty per cent. Per-user pricing against a workforce that moves with project cycles produces true-ups priced against a peak that no longer exists.
- Decide whether message history is an asset or a liability. If retention policy says eighteen months, you are paying to migrate content you are obliged to delete. That conversation belongs to legal, not to IT.
- Negotiate at the moment the alternative is credible. Price moves at renewal, when there is a real option on the table and time to execute it. It does not move in month four of a three-year term.
- Give collaboration licensing one owner across all entities. Three separate agreements with three renewal dates and three discount structures is the normal state of affairs and the single most reliable source of overspend.
The Regional Angle
Part of the bundle's value does not travel. Voice and video calling over the internet remains restricted to licensed operators in the UAE, and the consumer Skype service was blocked here at the end of last year. Enterprise calling is treated differently from consumer calling, and the answer depends on the entity, the licence and how the traffic is classified, but the instruction is the same for everyone: test calling and conferencing in-country, on your own network, before signing a business case that depends on them. Saudi Arabia relaxed its position on calling within messaging applications in September last year; the UAE's position is narrower. A consolidation case that counts conferencing minutes as a saving needs first to establish that the minutes are available. Licensing here is bought through regional resellers on enterprise agreements, negotiated per entity, with discount structures that differ across the same group. A holding company with mainland and free-zone entities routinely buys the same products three times on three renewal dates, which means nobody sees the total and consolidation savings evaporate into agreements that were never compared. The first deliverable of a licensing review in this region is usually a single sheet showing what the group actually buys. Workforce composition changes the maths in a way imported benchmarks miss. Contracting, hospitality, logistics and retail businesses here carry large deskless populations where only a minority of staff hold an account at all, which makes the firstline worker licence tier far more consequential than headcount suggests and makes the 300-user free cap map onto office headcount rather than total employment. A three-thousand-person contractor may sit comfortably inside a free tier. Set against that, per-user licensing runs against a workforce that expands and contracts with project awards and visa timing, so annual commitments made at a peak are paid for long after the peak has gone. And the genuine zero-cost incumbent for cross-company coordination is still WhatsApp, which no licensing decision touches and where a good deal of the work being licensed for actually happens.
The objection worth taking seriously
The objection is that free tiers do not move enterprise buyers, and the evidence for it is decent. Almost every organisation paying for Slack today already has Office 365 and bought Slack anyway. They did it because teams asked for it, and because the difference between a tool people reach for and a tool people are assigned is worth more than the licence. A free tier capped at 300 users does nothing to a four-thousand-seat deployment where the product is liked. Bundling has lost before, repeatedly, against products people actively prefer. The harder version is that the effect described here is a price ceiling rather than displacement, and a price ceiling is a slow problem. It surfaces three or four renewals out, in discounting and in accounts that stop growing, and nobody can point to a number this year. Making strategy out of it now is forecasting dressed as analysis. Both are fair, and neither changes what a buyer should do. The position that survives either outcome is the same: decide on the basis of what the tools do for the work, put the migration cost on the page next to the licence saving, and negotiate as though the alternative were credible. It is credible. That is the only part of this that is not a forecast.
Common Questions
Does the free Teams tier genuinely replace a paid messaging workspace?
For an organisation under 300 people whose requirements are messaging, search, file sharing, calls and guest access, it covers the ground. Where it does not is administration and compliance: limited administrative control, no retention or eDiscovery capability worth the name, and no service level commitment. A small regulated business hits that wall quickly. A small unregulated one may never notice.
We already pay for Office 365. Is Teams actually free for us?
The licence is. The deployment is not. Channel structure, governance, guest access policy, retention settings and the training that stops it becoming an unsearchable sprawl are real costs, and organisations that treat Teams as switched-on-already tend to pay them twice, once informally and again during the clean-up. Budget for the rollout even though the licence line reads zero.
What should we budget to move off an incumbent platform?
Plan on migration costing more than a year of the licence saving for anything above a few hundred users, with the variance driven almost entirely by integrations and by whether message history has to come along. Organisations that decide history can be archived rather than migrated tend to cut the project roughly in half.
What should we expect over the next twelve months?
The free tier reads as aimed at the small and mid-sized segment rather than at large incumbent deployments, so expect the visible effect in organisations that never bought anything, not in defections. Expect the bundle to keep absorbing adjacent categories, since Microsoft has already said Skype for Business Online capability will move into Teams over time, and each absorption widens the marginal-cost gap. Expect discounting on standalone collaboration tools to become more visible at renewal even where nobody switches. Expect the deskless and firstline licence category to grow faster than the knowledge-worker one, particularly in this region. And expect at least one organisation you know to consolidate, meet the integration bill, and quietly run both for another year.
Collaboration Licensing Review — we put the marginal cost, the migration bill and the renewal calendar for every entity on one page, so the decision gets made on arithmetic rather than on whichever vendor is in the room.