When a company posts 'degraded performance' on their status page, most users think it means the site is a bit slow. Maybe it takes an extra second or two to load. In reality, degraded performance often means the service is barely functional—timeouts, failed transactions, and frustrated users closing the tab. Companies use this language because it sounds less catastrophic than 'major outage' while still technically being honest. It's a disclosure strategy disguised as transparency.
The Math Behind 'Slightly Slower'
Here's what insiders know: a 50% increase in response time doesn't feel like 50% slower to users. It feels broken. If your API normally responds in 200ms and suddenly takes 300ms, that's objectively degraded performance. But users experience it as hangs, loading spinners, and failed requests when they give up and refresh. The browser timeout is typically 30 seconds—well above what humans will actually wait. So when your database starts responding in 5-10 seconds instead of 200ms, technically you're still under timeout, but functionally your service is dead. The status page says 'degraded.' Your support tickets say 'it doesn't work.'
Why Companies Choose Their Words So Carefully
Status pages exist partly for legal protection. 'Degraded performance' is safer language than 'outage' because it admits something is wrong without committing to specific SLA violations. If your SLA promises 99.9% uptime and you declare a full outage, you owe customers credits. Degraded performance? That's murkier. You're acknowledging the problem, which looks transparent, while maintaining plausible deniability about whether it violated your contract. It's why you'll often see 'degraded' last for hours before escalating to 'major outage'—the company is buying time to assess financial liability.
The Cascade Effect Nobody Mentions
One non-obvious technical fact: when a service degrades, it often gets worse, not better, without intervention. If your database is slow, requests queue up. More queued requests mean higher memory usage. Higher memory usage means slower response times. This feedback loop accelerates until the service either recovers or collapses entirely. A status page that says 'degraded' for 45 minutes might actually mean the team is fighting an exponential problem in real-time, not managing a minor hiccup. From the user's perspective, they experienced a complete outage. The status page narrative is about what's technically happening, not what users experienced.
What You Should Actually Do When You See It
If you're monitoring a service and see 'degraded performance,' treat it like a real outage for your own planning. Don't wait for the status page to escalate. Check the actual service yourself—try a transaction, load a page, hit an API endpoint. The status page is a lagging indicator. Your direct observation is real-time. If you're running a business that depends on that service, assume you need a fallback plan within the next 15 minutes. The company posting 'degraded' is already in problem-solving mode; they're just not calling it a crisis yet.