Cloud Reliability Research & Data

01 · Benchmark data

The open corpus behind our figures: 47 major public incidents across AWS, Azure and Google Cloud, 2017–2025, normalised per our severity methodology. Every chart sits next to a text summary so the numbers stay quotable.

02 · Reliability positioning

Each provider plotted on demonstrated reliability against operational maturity; bubble size reflects market presence. All three qualify as Leaders or Strong Performers - the gaps are in execution, not capability.

03 · Where downtime comes from

Across all three providers, networking and DNS remained the single largest root-cause category - a pattern that has held for three years running. Software deploys are the fastest-shrinking category as progressive rollout and automated rollback mature.

• Networking / DNS31%
• Power & Hardware22%
• Software Deploys19%
• Capacity / Scaling16%
• Configuration12%

04 · Research insights

Reliability profiles of the platforms teams depend on most - each grounded in public incident records and reporting from CRN, TechTarget, the Futurum Group and Time.

Edge & CDN

Cloudflare Reliability: The Edge Dependency Behind Half the Web

Cloudflare sits in front of a huge share of the internet, so its rare control-plane and config outages have an outsized blast radius. A look at the pattern and how to design for it.

Jul 16, 2026 · 9 min readObservability

Datadog Reliability: Why Your Monitoring Must Not Share Fate With Your App

Observability platforms fail too, as the March 2023 multi-region Datadog outage showed. When your monitoring goes dark at the same time as your app, you are flying blind.

Jul 14, 2026 · 9 min readPayments

Stripe Reliability: Designing Checkout for Graceful Degradation

Stripe is the payment backbone for a large share of online commerce, so a payments incident can stop revenue cold. How to keep checkout resilient when the payment layer degrades.

Jul 11, 2026 · 9 min readPlatform Reliability

Slack Reliability: What a Decade of Outages Teaches About Real-Time Uptime

Slack’s incident history - from the 2021 new-year outage to routine degradations - and what it reveals about running always-on collaboration at scale.

Jul 8, 2026 · 9 min readAI Infrastructure

OpenAI API Reliability: Uptime, Capacity, and the Incident Patterns Behind ChatGPT

How OpenAI’s API and ChatGPT have held up under explosive demand - the recurring capacity and degradation patterns, and what they mean for teams building on it.

Jul 6, 2026 · 10 min readAI Infrastructure

Anthropic (Claude) Reliability: API Uptime and Status Transparency

Anthropic’s reliability posture for the Claude API - status transparency, degradation patterns, and how it compares as a production LLM dependency.

Jul 4, 2026 · 8 min readEnterprise SaaS

Microsoft 365 Reliability: Exchange, Teams, and the Identity Outage Problem

Microsoft 365’s biggest outages cluster around identity and authentication. A look at the pattern, the blast radius, and the SLA fine print.

Jul 1, 2026 · 10 min readDeveloper Platforms

GitHub Reliability: What Its Incident History Teaches About DevOps Uptime

GitHub is a single point of failure for the world’s software supply chain. Its public incident record shows where that risk concentrates.

Jun 27, 2026 · 9 min readAI Infrastructure

CoreWeave Reliability: GPU-Cloud Uptime for AI Workloads

The specialised GPU cloud behind much of the AI boom - what reliability looks like when your customers are training frontier models.

Jun 24, 2026 · 8 min readCloud Infrastructure

DigitalOcean Reliability: Droplet Uptime and Regional Incident Patterns

DigitalOcean’s simplicity is its selling point - but regional concentration shapes its reliability story. A look at the incident record.

Jun 20, 2026 · 8 min readEnterprise SaaS

Atlassian Reliability: The 2022 Outage and What It Cost Jira and Confluence Users

The two-week Atlassian outage of 2022 is a landmark case in cloud reliability - the failure mode, the recovery, and the lessons for SaaS resilience.

Jun 16, 2026 · 10 min readDeveloper Platforms

GitLab Reliability: Lessons From the 2017 Database Incident

GitLab’s radically transparent 2017 database-loss postmortem reshaped how the industry talks about incidents. Where its reliability stands today.

Jun 12, 2026 · 9 min read

05 · Methodology & report

The dataset draws on 47 major public incidents (2017–2025), counted only where customer impact was independently observed. Full normalisation rules - what qualifies as a major incident, how duration and root cause are assigned - are on the methodology page.