Yes, VoIP is reliable, provided your provider’s network can keep calls moving when one of its sites goes offline. Your office needs a backup plan for its own internet and power too, since a provider can’t fix a dead modem from a data center three states away. You can check both before you sign, and you don’t need to be an IT person to do it.
Five nines (99.999% uptime) shows up in a lot of business phone ads, so the number by itself won’t separate one provider from another. What you want to know is what counts as downtime, what the provider owes you when it misses, and whether it shows its track record in public. Three places give you those answers: the service level agreement (SLA), the provider’s status page, and one question for the sales rep. How many of your sites handle live calls at the same time?
What the Nines Allow
Each nine you add cuts the downtime a provider is allowed by a factor of ten.
| Uptime | Downtime per year | Downtime per 30-day month |
| 99.9% | 8.76 hours | 43.2 minutes |
| 99.99% | 52.6 minutes | 4.32 minutes |
| 99.999% | 5.26 minutes | 26 seconds |
Twenty-six seconds a month assumes the provider counts every second of every interruption, and the SLA is where you find out whether it does. The two SLAs below both skip short interruptions, and both leave out anything that starts on your side of the connection.
What an Uptime SLA Covers and What It Pays
The SLA is the contract behind the uptime claim. It’s where a provider spells out what counts as downtime, what doesn’t, how you file for a credit, and what that credit is worth. Two published examples show how much those terms can vary.
Take Intermedia’s SLA for Unite and Cloud Voice, which targets 99.999% a month. An interruption under five minutes doesn’t count, and neither does scheduled maintenance (up to 60 minutes a month) or trouble with internet availability or IP transit. If you want a credit, you have 72 hours to open a support ticket, and the written request goes in the following month. Availability anywhere between 99.0% and 99.999% gets you 3% of that month’s fee for the affected service. Credits top out at 50%, and the agreement says the credit is your only remedy.
RingCentral’s SLA for MiCloud and Sky (both formerly Mitel) draws the line somewhere else. Downtime only counts once you’ve been unable to receive inbound calls for 30 minutes straight, and your credit request has to reach them in writing within 10 calendar days. A month between 99.50% and 99.98% earns 5% back. There’s one term in it worth asking every provider for, though: if uptime drops below 99% two months in a row, or three times in six months, you can terminate. (This SLA covers MiCloud and Sky only. RingCentral’s main product isn’t under it.)
Do the math on a $1,000 monthly bill. A 3% credit is $30, and it doesn’t cover the calls that went to voicemail or the customer who hung up and called someone else. So on a sales call, spend less time on the uptime percentage and more on the termination clause, since that’s what lets you walk away from a provider that keeps missing. For more contract fine print to check before you sign, read The Hidden Costs in Your Business Phone Bill (and How to Read Them).
How a Network Stays Up: Points of Presence and Failover
Points of presence (PoPs) are the sites where a provider’s equipment hands your calls off to the rest of the phone network. How a provider sets them up decides what happens when one of them fails.
Some providers keep one primary site and a standby backup. The backup waits until the primary goes down, and switching over can take time or need a person to make the call. Other providers run several sites live at once. Each one is already handling calls, so when one drops out, the others pick up its traffic automatically. That handoff is called failover.
CallHarbor uses the second setup. It runs three fully active points of presence in Detroit, Atlanta, and Phoenix with automatic failover, so if one or two go offline, traffic moves to the site that’s left. In the 13 years since CallHarbor moved to that model, all three have never been down at the same time. It also runs its own hardware and its own network in diverse data centers, which means fewer outside vendors are involved when something needs fixing.
Reliability is the second of CallHarbor’s three pillars (customer service is first, value is third). Here’s how founder Frank Kadaf describes the work behind it:
“We spent a lot of time, a lot of effort, and a lot of money building a platform that always works and is reliable, by reducing dependencies on third parties,” says Frank Kadaf, Founder & CEO of CallHarbor.
What Failover Looked Like in a Real Incident
CallHarbor’s own status page has an example from September 4, 2023. A problem at the Detroit PoP took down eFax, call recording, and provisioning for services running out of Detroit, and the first update on the page said calls were working normally. By the next day, recording and provisioning were running from other points of presence. eFax came back on September 6 after a database restore from backups.
Calling stayed up. When you read any provider’s incident history, that’s what you’re hoping to find: a problem that stays inside one site or one feature while people keep making and taking calls.
How to Read a Provider’s Status Page
Ask any provider you’re considering for its status page link before you sign, and give yourself ten minutes to look around. Start with the dates. A history that only goes back a few weeks won’t teach you much, and CallHarbor’s at status.callharbor.com goes back more than three years.
Look at how the page is organized. CallHarbor lists Detroit, Atlanta, and Phoenix as separate groups, each with its own components like inbound calling, outbound calling, and phone registration, so you can tell at a glance whether a problem hit one site or all of them.
Maintenance notices are worth reading, too. A good one tells you when the work happens and which sites will keep taking calls in the meantime. When CallHarbor scheduled network maintenance at its Atlanta PoP on February 13, 2025, the notice said services would keep running from Detroit and Phoenix.
Check whether the provider logs problems with add-ons like fax, call recording, and the mobile app, or only with core calling. CallHarbor’s page includes add-on incidents such as eFax. A provider that only posts its biggest incidents is giving you a partial record.
Last, subscribe. CallHarbor’s page sends updates by email, Slack, Microsoft Teams, RSS, and other channels. During an incident, you’ll also want to know who answers when you call support, and Live Human Support for VoIP vs. Tiered Call Queues: What First-Call Resolution Saves You covers what to ask.
Your Side of the Connection
Neither SLA above covers your own internet, and three live data centers can’t help a desk phone in an office that’s lost its connection. That part is on you.
Start with a second internet connection, either from a different provider or an LTE backup, plus a router that switches over on its own. Power is the next gap. A lot of desk phones get their power from a PoE (power over Ethernet) switch, so a power outage can take them down even if the internet is fine. A battery backup on the router and the PoE switch keeps them going through short outages.
Set up call forwarding rules that send calls to cell phones or another location when desk phones drop off, and test them before you need them. CallHarbor customers can also use the CallHarborUC mobile app to make and take business calls from their cell phones while the office is down.
Some outages have nothing to do with your phone provider. In January 2026, Verizon had a nationwide outage, and calls from AT&T and T-Mobile customers to Verizon customers failed too. Afterward, Verizon offered affected customers an account credit, which is the same kind of remedy an SLA gives you.
Seven Checks Before You Sign
- Ask for the SLA document itself, along with the uptime number on the website.
- Find the shortest outage that counts and the deadline to file for a credit.
- Read the exclusions, including maintenance windows and your own internet.
- Ask whether you can leave the contract after repeated misses.
- Open the status page. How far back does it go, and is each site listed?
- Ask how many sites handle live calls at once and what happens to calls when one goes offline.
- Set up a backup path for your own internet and power.
Frequently Asked Questions
Is VoIP more reliable than a landline?
It depends on what goes wrong. An old copper landline gets power from the phone line, so it keeps working through a power outage, and VoIP desk phones need a battery backup to do the same. When a provider’s site fails, VoIP has the edge, because a provider with several live data centers can route around the problem and your forwarding rules can push calls to cell phones. Add battery backup and a second internet connection, and VoIP covers both cases.
What are the most common VoIP problems?
Choppy audio, dropped calls, and one-way audio (you can hear them, they can’t hear you) come up a lot. The cause is often the office network: not enough bandwidth, congestion at busy times, or a router that isn’t set to prioritize voice traffic. Problems on the provider’s network should show up on its status page, which is a quick way to tell which side you’re dealing with. If call quality issues keep coming back, ask your provider to look at your network setup and your router’s quality of service (QoS) settings.
Is 99.99% uptime for a VoIP provider good?
It’s a solid target, assuming the SLA counts downtime fairly. 99.99% allows 52.6 minutes of downtime a year (4.32 minutes in a 30-day month). Many voice providers advertise 99.999%, which drops the yearly figure to 5.26 minutes. Either way, compare the SLAs before you compare the percentages.
Does a VoIP uptime SLA pay for missed calls?
No. Both SLAs in this post pay a credit worth a percentage of one month’s fee, and Intermedia’s names the credit as your sole remedy. Lost sales and missed calls aren’t part of it, so a provider’s network design and incident history tell you more about your risk than its credit table does.
How can I check a VoIP provider’s outage history?
Ask for the status page URL. From there, check how far back the history goes, whether each data center has its own listing, and how long past incidents took to resolve. CallHarbor’s is public at status.callharbor.com.
Check Your Current Provider First
Before you shop around, pull up the SLA and status page for the provider you have now and go through the seven checks. If short outages don’t count, the credit is a few percent of one month’s bill, and the status page only goes back a few weeks, you’ve got your questions for renewal.
CallHarbor’s record is public at status.callharbor.com. If you want to hear how the network handles a site going down, call 833-363-VOIP and talk to someone on the team.