Microsoft 365 services carry a monthly uptime commitment. Below 99.9% the typical credit is 25%, below 99% it's 50%, and below 95% it's 100%.
Uptime is counted in user-minutes, not clock time. "44 minutes down" only qualifies if every user was fully down for that long.
Credits are not automatic and cover only the affected service's fee, as a discount on a future invoice. You or your reseller must file a claim with the incident ID and evidence.
For the November 2026 outage (MO1550821, EX1550543), record your impact now and file after Microsoft closes the incident. Reseller (CSP) claims are due within one month after the end of the billing month.
Yes, you may be owed a service credit for the 12–13 November 2026 Microsoft 365 outage, but probably less than social media suggests. Microsoft's credits start when a service's monthly uptime falls below 99.9%. That uptime is weighted by how many users were affected and for how long. The credit is a percentage of that one service's fee, it comes off a future invoice, and only an admin or reseller claim triggers it.
How the uptime percentage is calculated
Microsoft measures uptime per tenant per calendar month, in user-minutes: total user-minutes in the month, minus the user-minutes of downtime, divided by the total (Microsoft Learn; worked example). November has 30 days, which is 43,200 minutes per user. So if everyone in the tenant is completely down:
Below 99.9% (25% credit): more than about 43 minutes.
Below 99% (50% credit): more than about 7.2 hours.
Below 95% (100% credit): more than about 36 hours.
Those thresholds stretch when only part of the company is affected. Take a 200-user tenant where 100 users lost Outlook for 10 hours. That's 60,000 downtime user-minutes out of 8,640,000, about 0.7%. Monthly uptime is roughly 99.3%, which is the 25% tier, not 50%. With all 200 users down for the same 10 hours, uptime drops to about 98.6% and you reach 50%.
"Down" also has a definition, and it differs per service. Microsoft's own example: for Teams, downtime is any period when users can't see presence, chat, or start online meetings. During this outage Microsoft said core Teams messaging and meetings largely kept working, so Teams may not count as down for many tenants even if calendar scheduling failed. Check each service's definition in the Service Level Agreement for Online Services, which holds the binding terms and exclusions.
What the credit is actually worth
Only the affected service: credits are prorated by service. An Exchange Online outage earns credit for the Exchange part of your licence, not your whole Microsoft 365 bill (Partner Center).
Not a refund: it is applied to a future invoice once Microsoft approves the claim.
Advisories rarely count: incidents posted only as advisories (the tenant "might" be affected) typically aren't eligible. MO1550821 and EX1550543 were posted as incidents.
Exclusions apply: the SLA excludes causes such as your own or third-party hardware and software. Read the current exclusions before you file.
For most mid-sized tenants the amount is modest. The record of impact is usually worth more than the money: it is the evidence your auditor will ask for under third-party risk management.
How to claim
Record the impact now. Note when users lost access, when it came back, how many and which users were affected, and what you tried. Microsoft asks for the downtime, the affected users and your attempts to resolve it.
Wait for the incident to close. Microsoft's post-incident report in Service health gives the official timeline to reference.
Bought directly from Microsoft: file a service request from the Microsoft 365 admin center (Support → New service request). Ask for an SLA credit and quote the incident IDs, here MO1550821 and EX1550543, plus your impact details.
Bought through a reseller (CSP): ask your partner. Only CSP direct and indirect providers can file, in Partner Center, within one month after the end of the billing month in which the incident happened. They need your tenant ID (GUID), the incident ID, and proof of impact sent from an email address on your tenant's own domain (Partner Center). Some distributors file collectively for large outages, so ask whether yours does.
There is no need to rush a claim on the day of the outage. There is a need to capture the facts while people still remember them.
Use it for NIS2 and DORA evidence too
The same notes belong in your incident register. DORA requires financial entities to record all ICT-related incidents, including those caused by a provider, and NIS2 asks you to judge whether the disruption was significant. Our outage write-up covers that side, and the incident clock gives you the reporting deadlines if it turns out to be significant.
Questions
Does Microsoft give SLA credits automatically after an outage?
No. Your admin, or your CSP reseller, has to submit a claim with the incident ID and details of the downtime and affected users. Microsoft reviews it and, if approved, applies the credit to a future invoice.
How long does a Microsoft 365 outage have to last to get a credit?
Long enough to push that service's monthly uptime below 99.9%, measured in user-minutes. With every user fully down that's about 43 minutes in a 30-day month; if only part of your users were affected, it takes proportionally longer.
What is the deadline to claim an SLA credit for Microsoft 365?
For subscriptions bought through a CSP partner, the claim is due within one month after the end of the billing month in which the incident occurred. If you bought directly, check the SLA for Online Services and file once the incident is closed.
Want your Microsoft 365 risk documented properly?
We review how your tenant and processes hold up when Microsoft has a bad day: continuity, break-glass access, out-of-band communication, and the third-party risk evidence NIS2 and DORA auditors ask for.