On September 16th, Salesforce — a $300 billion company with tens of thousands of enterprise customers — went dark. Login failures. API errors. Dashboards that wouldn’t load. Sales teams sitting idle during Dreamforce, the company’s flagship annual conference.
The cause wasn’t a cyberattack. No ransomware, no breach. A single internal login service stalled. That stall cascaded into an external dependency failure, which then hammered a core system component until it ran out of resources and brought everything else down with it.
That’s the pattern behind most major cloud outages. Not dramatic hacks. A quiet failure in one place that the rest of the system wasn’t built to contain.
The impact hit customers across multiple regions simultaneously. Sales pipelines inaccessible. Customer service teams unable to pull records. Automated workflows and scheduled jobs stopped cold. Engineers deployed a validated fix fleetwide — but some customer instances needed manual restarts, and scheduled jobs continued failing well into the recovery window.
What Cloud Dependency Actually Costs
Most businesses haven’t done the math on what an hour of platform downtime costs them. Take a 10-person sales team averaging $100,000 per year. That’s roughly $480 an hour in idle payroll, before you count lost pipeline, stalled customer follow-ups, or the downstream effects on any process that depends on CRM data.
The trade-off with SaaS platforms is real. You get enterprise-grade capability without running your own infrastructure. What you give up is control. When the vendor’s systems fail, your operations fail with them — and you’re in a queue waiting for their engineers to fix it.
“Restored service” also doesn’t mean “back to normal.” Salesforce returned most customers to baseline by roughly 6:56 AM Eastern. But queued scheduled jobs, automated sequences, and workflow triggers that fired during the outage don’t automatically retry correctly. Recovery takes time after the platform is back up.
Three Practical Steps Before the Next Outage
First, map your critical SaaS dependencies and document what breaks when each one goes down. Most companies have no runbook for “Salesforce is unavailable for four hours.” Write it before you need it.
Second, build offline fallback procedures. What does your sales team do when the CRM is inaccessible? The answer can’t be “wait.” Even a shared spreadsheet process beats paralysis.
Third, subscribe to your vendors’ status pages and set up alerts. Salesforce publishes a public status feed. Knowing about an outage 20 minutes faster means you start managing it 20 minutes earlier — that’s 20 minutes less chaos.
Cloud platforms are generally more reliable than anything most businesses would run themselves. But “generally reliable” is not the same as “always available.” The gap between those two things is what the Salesforce outage just made visible.
Want to explore how to build more resilient operations around your cloud platforms? Let’s talk.