We sell bonding, and we’ll tell you when not to buy it. If your site fits the profile below, a simpler design serves you better — and you’ll trust the rest of this site more for knowing that.
Failover really is enough when…
- Your primary circuit is genuinely reliable (business fiber with a real SLA, not best-effort coax) and a 30–60 second cutover once or twice a year costs you annoyance, not money.
- Nothing on site holds a session that can’t die: web, email, and cloud apps reconnect on their own. (Card terminals, VoIP calls, live video, and RDP sessions do not — that’s when you need hot-failover or bonding.)
- Budget is the constraint: a dual-WAN router in priority-failover mode with a 5G SIM standing by costs less than a bonded multi-carrier build, and it eliminates the “internet is down, everyone go home” day. That alone is most of the value for many small sites.
A simple honesty checklist
| Question | If yes → |
|---|---|
| Does a dropped call / declined card / frozen stream cost you money or reputation? | Hot-failover at minimum; bonding if it happens under load |
| Is any single carrier weak at the site? | Multi-carrier — bond what exists |
| Do you upload big files on a deadline (video, imaging, surveys)? | Bonding — aggregate the uplinks |
| None of the above? | Dual-WAN failover + monitoring. Save the difference. |
What “good failover” still requires
If you go the simple route, do it properly: health checks against real destinations (not just the gateway), alerts when the backup activates (a failover nobody notices is a bill nobody expects), a monthly test, and a data plan that won’t bill-shock you the week the fiber gets cut. Our User Guide covers each step, and any Peplink you buy for failover today can turn on full SpeedFusion bonding later — the engine ships in the box, no license expiry.
Not sure which side of the line you’re on? The kit selector asks exactly these questions, or talk to an engineer — we’ll tell you if failover is enough.