What Actually Happens When Your Primary ISP Goes Down
An ISP outage doesn't ask permission. Here's what automatic failover actually does when the primary link drops, and how to tell if it's working.
Every property with guest WiFi eventually has the same afternoon: the internet drops, the front desk starts fielding complaints, and someone spends twenty minutes on hold with the ISP trying to figure out whether it’s their fault or the fault of the guy digging up a cable three blocks away. By the time anyone has an answer, you’ve already lost the goodwill of everyone who was mid-video-call, mid-checkout, or mid-upload when it happened.
The fix for this isn’t “get a better ISP.” Every ISP goes down eventually — that’s not a failure of vendor selection, it’s how internet infrastructure works. The fix is not depending on a single uplink in the first place.
Primary and backup, actually switching
Failover means running two internet connections into a location — a primary link and a backup — and automatically switching guest traffic to the backup the moment the primary drops, without anyone at the property doing anything. No one has to notice the outage first, log into a router, or call anyone. The switch happens because the system is continuously checking whether the primary link is actually up, not just assuming it is.
That last part is the detail that matters. A lot of “redundant” setups aren’t actually monitored — the backup line exists, but nothing is watching the primary closely enough to fail over quickly, so it just sits there unused until someone manually flips it during an outage that’s already been going on for an hour. Automatic failover only works if the health check behind it is real and frequent.
What “real” WAN health actually looks like
Live WAN health per location means latency, packet loss, current throughput, and an actual up/down history — not a single green “connected” icon that tells you nothing until it turns red. A dashboard that shows one status light is telling you the connection is either fully up or fully down; it isn’t telling you that packet loss has been climbing for the last twenty minutes, which is usually the actual warning sign before a full outage.
The up/down history matters just as much as the live number. When a guest or a manager asks “was the WiFi actually down yesterday afternoon,” a real down-duration timestamp — three hours and twelve minutes, starting at 2:14pm — is a very different conversation than “I think it might have blipped.” One is a fact you can act on or dispute with the ISP; the other is a guess.
Why the backup link needs its own health check too
Failover is only as good as the backup connection it switches to. A backup line that’s never checked isn’t insurance, it’s an assumption. Real failover setups monitor both links — primary and backup — so that when the primary drops, the system already knows the backup is actually reachable before it routes guest traffic there. Switching to a backup line that’s also down doesn’t help anyone; it just moves the outage from unnoticed to actively confusing.
What this actually protects
Failover doesn’t touch the guest login experience, the branding on the portal, or which devices get access rules applied to them — those are separate layers. What it protects specifically is continuity: a guest who’s already connected and streaming, working, or checking out doesn’t notice the primary ISP dropped, because their traffic moved to the backup link before the connection actually died from their side.
For a hotel, that’s the difference between a guest who never knew there was an outage and a guest who’s calling the front desk asking why their video call cut out. For a co-working space, it’s the difference between a member losing an hour of billable work and not losing anything at all. Neither of those outcomes depends on which ISP you picked — they depend on whether the switch to backup happened automatically and fast enough to matter.
What to actually check
If you’re evaluating whether your current setup does this, the questions worth asking are specific ones. Is there a second physical uplink, or is “backup” just a phrase in a contract? Is the primary link’s health checked continuously, with real numbers — latency, packet loss, throughput — or just a binary status? When the primary does go down, does the switch happen automatically, or does someone have to notice and act? And afterward, can you actually pull up how long the outage lasted and when it started, or is that information already gone by the time someone asks?
None of these are exotic requirements. They’re the difference between a guest network that degrades gracefully during an ISP outage and one that just goes dark until a human intervenes — and on a property that depends on guest WiFi actually working, that difference is the whole point of having a backup connection at all.