A Microsoft 365 user cut off for the whole of Microsoft’s recorded outage window, 1,430 minutes on 12-13 November, ends November at about 96.7% uptime, by our arithmetic. Under Microsoft’s October 2026 Service Level Agreement, that is below the 99% line and above the 95% line: the band where the credit is 50% of the monthly fee.
The SLA counts downtime in user-minutes for each customer, so a tenant with only some users affected scores higher. By our arithmetic, with no other downtime in the month, a tenant falls below 99.9% if more than 3.0% of its users were cut off for the full window, and below 99% if more than 30.2% were. The recorded window alone cannot take any tenant below 95%.
Business Pill · Embedded Dependency
A short explainer on embedded dependency: what sits inside the systems you buy. It teaches the general idea only and does not discuss Workday or the outage.
The key insight: As we read it, the headline window and the credit are two different numbers: the incident window puts a fully cut-off user at about 96.7%, but each tenant’s uptime is weighted by how many of its users were affected, so the same outage can sit in the 50% band for one customer and outside any band for another.
Microsoft’s Service Level Agreement for Microsoft Online Services, October 2026 edition, defines Exchange Online downtime as “Any period of time when users are unable to send or receive email with Outlook Web Access”. It adds: “There is no Scheduled Downtime for this service”. The credit table pays 25% of the fee below 99.9% monthly uptime, 50% below 99% and 100% below 95%; Microsoft Teams and Microsoft 365 Apps carry the same table.
Downtime is “measured in user-minutes”: the length of each incident “multiplied by the number of users impacted by that Incident”, against the month’s minutes times all users. Microsoft 365 Copilot is measured differently, by the share of valid prompts that get an unsuccessful response, and its first credit line is 99.5% rather than 99.9%.
Minutes of downtime each credit line of Microsoft’s SLA allows in a 30-day month, against the recorded 1,430-minute incident window and the 2,175-minute span of Microsoft’s public posts. All figures are our arithmetic from the October 2026 SLA and Microsoft’s record.
The 96.7% Count
Microsoft’s final service-health entry, as republished by Cloud Admin Hub, gives “Start time: Thursday, November 12, 2026, at 4:20 PM UTC” and “End time: Friday, November 13, 2026, at 4:10 PM UTC”: 1,430 minutes, by our arithmetic. November has 43,200 minutes, so 99.9% leaves 43.2 minutes of downtime, 99% leaves 432 and 95% leaves 2,160.
A user out for all 1,430 minutes scores (43,200 − 1,430) / 43,200, about 96.7%. Reaching the 95% line would take 730 more minutes, about 12 hours, by our arithmetic. Microsoft’s public posts about the incident spanned 36 hours 15 minutes, or 2,175 minutes, as we counted in our root-cause piece; a user cut off for that whole span would sit at about 94.97%.
What a Credit Covers
Credits are based on “the total fees actually paid by you for a Service” in the period, and the SLA says they “will not, under any circumstance, exceed your monthly service fees” for that service. It also says: “Service Credits will not be awarded to compensate for any other forms of loss, including but not limited to lost revenue, operational costs, or any indirect losses experienced by you or the end-users”.
Credits are not automatic. The SLA says the customer must submit a claim “including all information necessary for Microsoft to validate the claim”, and that claims are processed “typically within forty-five (45) days of receipt of the claim”. Its example says an incident on 15 February must be claimed by 31 March; for a November incident, that points to 31 December, by our reading.
The Structural Read
As we read the SLA, the remedy is narrow by design: credits are capped at the monthly fee for the affected service and the SLA excludes lost revenue and operational costs, so a credit returns part of a subscription, not the cost of a lost day.
As we read the claim rules, the burden sits with the customer: the claim must carry the time, duration, affected users and errors, and Microsoft’s own outage posts are not a confirmation of a missed service level.
Microsoft, Service Level Agreement for Microsoft Online Services, October 2026
“Service Credits are your sole and exclusive remedy for any performance or availability issues for any Service under the Agreement and this SLA”
Three Implications
FOR IT ADMINS The SLA asks for the number and location of affected users and the time and duration of downtime with each claim.
FOR FINANCE TEAMS A November incident points to a 31 December claim deadline under the SLA’s own example, by our reading.
FOR COPILOT BUYERS Copilot is measured by unsuccessful responses to valid prompts, with a 99.5% first credit line, not by user-minutes.
The framework says: “The value proposition shifts from features to integration depth.”
As we read it, the SLA shows the price of that depth from the contract side: the remedy for a day without email is capped at a share of one month’s fee. The framework predates this outage and does not discuss it.
Share of a tenant’s users cut off for the full 1,430-minute window needed to fall below each credit line, with no other downtime in the month. Our arithmetic from Microsoft’s user-minutes formula.
What Is Not Established
Microsoft’s full post-incident report is not out yet. How many users at each customer were cut off, and for how long, is not known, so no customer’s actual uptime or credit can be read from the incident window. The SLA says Microsoft’s outage communications “are not a confirmation of missed Service Levels or Service Credits eligibility”. Enterprise agreements can carry different terms. We did not contact Microsoft for this piece.
This Week In The AI Supercycle
AI Lost Its Central Banker, the Lab Oligopoly, and the Future of Open Source
This week’s issue: the Nvidia leadership change, read as the AI build-out losing its central banker: who now finances the GPU landlords, backstops the data centres and sponsors open models, and why two labs are becoming the bank.
By our arithmetic, a Microsoft 365 user cut off for the full 1,430-minute window ends November at about 96.7% uptime, in the 50% credit band of Microsoft’s SLA; a tenant needed more than 3.0% of its users out all window to fall below 99.9%. The outage was one of three we counted in 3 Outages, 46 Hours in Two Days: Microsoft, Workday, the FAA.
A note on sourcing. The SLA text, credit table, claim rules and fee cap come from Microsoft’s Service Level Agreement for Microsoft Online Services, October 2026 edition, downloaded from Microsoft’s licensing site. The incident window comes from Microsoft’s final service-health entry as republished by Cloud Admin Hub. Uptime percentages, minute thresholds, user shares and the claim date are our arithmetic and reading; they apply the published terms to the recorded window and are not any customer’s actual uptime. We did not contact Microsoft.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
