Outage Record

Microsoft has had more big outages than anyone else since 2020

Every big outage since 1 January 2020 that the provider admitted in public, that hit customers in more than one country (more than one state, province or region for a telco) and that lasted an hour or more. 207 outages across 80 providers so far. Each one links to its source.

207
big outages on record since 1 January 2020
27
of them were Microsoft
About 249 hours
Microsoft down in total, from published durations
61.4 hours
the longest: KDDI au, 2 July 2022 433

The outage leaderboard

Ranked by number of outages, then total hours. Filter by type. Hours add up published durations only.

Microsoft leads on number of outages

Big outages since 1 January 2020 under the rule on this page.

Microsoft
27
Google
14
Cloudflare
10
Meta
8
Atlassian
6
Slack
5
OpenAI
5
Zoom
5
X (Twitter)
5
Optus
4
Apple
4
Verizon
4
Deutsche Telekom
4
Three UK
4
Salesforce
3
AWS
3
GitHub
3
AT&T
3
PayPal
3
NTT Docomo
3

Top 20 shown. Telcos in grey. The table above has all 80.

Microsoft also leads on hours down

Total hours from each provider's published durations. Outages with no published length add nothing here.

Microsoft
248.6 h (2 with no published length)
Google
76.6 h
KDDI au
64.3 h
Cloudflare
59.8 h (1 with no published length)
Salesforce
56.7 h
Atlassian
41.8 h (1 with no published length)
Slack
40.1 h
Optus
32 h
Canva
32 h
OpenAI
26.9 h (1 with no published length)
Rogers
24 h (1 with no published length)
AWS
22 h (1 with no published length)
GitHub
20 h
Meta
17.5 h (2 with no published length)
Orange France
16.4 h
Zoom
13.6 h
AT&T
12 h (2 with no published length)
T-Mobile
12 h (1 with no published length)
PayPal
11.2 h
Apple
11 h

Top 20 shown. Telcos in grey. The table above has all 80.

The ten longest outages on record here

Single outages, longest first.

How the board is counted

An outage counts when a listed cloud or software vendor, or a national telco, acknowledged it in public, it hit customers in more than one country (more than one state, province or region of one country for a telco), and it lasted 60 minutes or more. The rule was written before any counting began, and earlier years are minimums because older incidents are harder to find.

A count is a count of incidents. Bigger providers run more services, so they have more chances to fail, and some publish incident reports more readily than others. Treat the board as a record of what happened, with that in mind.

Microsoft includes Azure and Microsoft 365. Google includes Google Cloud and Workspace. Hours come from the provider's own report where one exists, and from named news reports otherwise.

The board covers 25 cloud and software providers and 55 telcos in 22 countries. Every outage is in the full list with sources, and the methodology explains how each one was checked.

Questions about the leaderboard

Which company has had the most major outages?

Microsoft, with 27 of the 207 outages on this board since 1 January 2020. That counts Azure and Microsoft 365 together, because both are run by Microsoft.

What was the longest cloud outage since 2020?

KDDI au on 2 July 2022: Mobile voice and data, about 61.4 hours by the published account.

How often does Microsoft 365 go down?

Microsoft had 27 outages on this board since 2020 that hit more than one country for an hour or more. Shorter or single-country incidents happen more often and aren't counted here.

Who has had the most outages in 2026?

So far in 2026: Microsoft 6, Google 2, GitHub 2, Cloudflare 1, Meta 1, Atlassian 1, Slack 1, OpenAI 1, Apple 1, Verizon 1, Salesforce 1, AWS 1, Telstra 1, Singtel 1, One NZ 1, Orange France 1, T-Mobile 1, Jio 1, 2degrees 1, Telkom Indonesia 1, Zendesk 1, Vodafone Australia 1, Xero 1, AT&T Mexico 1, Movistar 1.

Keep working through the next one

Check the live status board when something stops working, and use the free outage playbook to set up a fallback before you need it.

Three places to go from here