Salesforce, the software that runs a huge chunk of American business, went down for thousands of users this week, and the timing could hardly be worse.
It's the middle of the quarter, the stretch when sales teams are scrambling to close deals before the calendar flips.
Instead of pipeline dashboards and account records, many users got error pages and the digital equivalent of a busy signal.
The outage tracker Downdetector lit up with reports from across the country, with complaints spiking in the morning hours before tapering off.
Salesforce later acknowledged service disruptions on its status page, saying it was investigating the issue.
For the uninitiated, that sounds like a tech hiccup.
For the companies that live inside this software, it's closer to the lights going out in the middle of a shift.
It's the system of record for sales, customer service, and marketing at tens of thousands of companies, from small outfits with a handful of seats to Fortune 500 giants.
When it wobbles, orders don't get logged, support tickets don't get answered, and commissions sit in limbo.
A few hours of downtime can ripple into days of cleanup.
The bigger question is why so much of American business depends on one company's servers behaving.
Cloud software meant companies could stop babysitting their own hardware and let someone else handle the plumbing.
That trade-off made sense on the spreadsheet.
It makes less sense when a single vendor's bad afternoon becomes your bad afternoon.
There's also a money angle most people miss.
Salaried reps still get paid while they wait.
Deals that were supposed to close this week might slip into next month, and depending on how the numbers fall, that can nudge quarterly revenue and the bonuses tied to it.
Nobody sends you a bill for an outage, but the bill shows up anyway.
For regular folks, this is mostly a reminder about the hidden fragility behind everyday transactions.
The rep who's supposed to call you back about your insurance quote, the customer service agent handling your refund, the utility scheduling your service visit — a lot of them are working inside tools like this.
When the tool breaks, your errand stalls too, and you never get an explanation beyond "our systems are down." If you run a small business and lean on a cloud platform, the practical takeaway is boring but real: know how to work offline for a few hours.
Keep key phone numbers somewhere that isn't the cloud.
Export critical contact lists periodically.
It's not glamorous, and most days you'll never need it, but the day you do, it beats staring at an error screen.
It's also fair to ask what customers are getting for the subscription fees.
Reliability promises in contracts often come with credits that amount to a rounding error on the bill, while the actual cost of the disruption lands on the customer.
That asymmetry is worth remembering the next time a vendor pitches "enterprise-grade" anything.
My take: outages happen, and no platform is immune, so I'm not ready to declare the cloud a mistake.
But the gap between how dependent businesses are on these tools and how little recourse they have when the tools fail keeps widening.
Final Thoughts
Until that changes, every "minor service disruption" is a reminder that someone else holds the switch.