An Internet outage doesn't have to mean that your customers hear nothing but a busy signal or an automated message. If you want to plan for VoIP reliability, you can’t just focus on the phone system. The entire chain is crucial: power supply, Internet access, network, SIP trunk, PBX, end devices, and the people who respond in the event of a failure. Only when these elements work together can your business remain reachable.
For small and medium-sized businesses, the goal is not the theoretically highest level of availability. A sensible solution is one whose protection is tailored to the business risk, remains affordable, and works without complicated steps in an emergency. A business with a single central phone number and just a few workstations needs a different approach than an organization with multiple locations, a customer hotline, and Microsoft Teams as its primary daily communication channel.
Planning for VoIP resilience starts with assessing business risk
The right question isn't, „How can we prevent every disruption?“ That would be virtually impossible from a business perspective. Instead, ask: What would be the consequences if our main number were unavailable for 30 minutes, four hours, or an entire workday?
For an architecture firm, temporarily redirecting calls to cell phones may be sufficient. For an on-call service, a medical office, or a company with a high volume of orders, even a brief interruption can lead to missed orders, security risks, and damage to the company’s reputation. This leads to clear objectives: How quickly should phone service be restored? Which functions must be available at all costs? And which ones can be temporarily suspended?
Also, document any dependencies. Does the cloud PBX run over a single Internet connection? Does the router also serve as a firewall, Wi-Fi hub, and SIP gateway? Is the phone system located in your own server room and connected to a single UPS? Questions like these reveal individual points of failure that often go unnoticed in day-to-day operations.
The Most Common Points of Failure in the VoIP Chain
VoIP is not inherently more vulnerable than traditional telephony. However, the dependencies are different. Whereas in the past a separate telephone cable ran all the way to the office, IP telephony typically uses the existing data infrastructure. This provides flexibility but requires careful planning.
Internet Access and Location Network
If the primary Internet connection fails, local IP phones typically lose their connection to the cloud PBX and the SIP trunk. For business-critical locations, a second, technically independent connection is therefore recommended. This can be a second landline via a different infrastructure or a cellular backup.
For many small and medium-sized businesses, a 4G or 5G fallback provides good protection against service interruptions. Its limitations lie in capacity, wireless coverage, and potential network congestion during an incident. It is often sufficient for a single location with ten concurrent calls. For a contact center or a large production site, however, bandwidth, prioritization, and the technical separation of access paths should be examined more closely.
The firewall is also an integral part of the design. It must reliably support the switch to the backup connection and handle SIP and RTP traffic in a controlled manner. An incorrectly configured SIP-ALG or missing Quality-of-Service rules can impair voice quality, even though the Internet connection is generally available.
Power Supply and Local Hardware
Without power, routers, switches, Wi-Fi, and IP phones won’t work. A UPS bridges short outages and enables an orderly shutdown during longer outages. It’s not just the rated power that matters. Check which devices actually need to be powered: firewalls, PoE switches, access points, local SBCs, and, if necessary, an on-premises PBX server.
A UPS is no substitute for an emergency power plan. For locations prone to prolonged power outages, a generator may be a sensible option, but only if its maintenance, fuel supply, and switchover time are realistically factored into the plan. It is often more cost-effective to automatically reroute the main phone number to external destinations during an outage rather than keeping all office workstations running.
Phone System, SIP Trunk, and Session Border Controller
Cloud PBX solutions reduce the overhead associated with on-premises servers and simplify use across multiple locations. However, a defined plan is still needed in case a location, a device, or the connection becomes unavailable. Therefore, it should be possible to forward calls to alternative destinations regardless of the local telephone system.
At local telephone systems The high availability of the PBX itself is particularly important. Depending on the size of the environment, redundant systems, secure configurations, and clearly documented recovery procedures may be appropriate. A backup that has never been tested is not a failover strategy.
A session border controller protects the boundary between the corporate network and the service provider’s network and can play a central role in more complex environments. It controls signaling and media streams, supports secure connections, and isolates internal systems from the public network. Whether a high-availability SBC is necessary depends on the number of locations, call volume, existing security requirements, and Teams integration.
Call forwarding is the most effective emergency mechanism
If the office is unreachable, calls must still be routed to someone. That’s why a preconfigured emergency call forwarding feature is an essential part of any VoIP solution. It can route calls to cell phone numbers, another office location, an external reception desk, or a temporary automated message.
Call forwarding should not be set up only after a disruption occurs. Determine in advance which numbers are affected, the order in which destinations will be called, and who is authorized to trigger the change. Be sure to take into account group numbers, IVR menus, queues, and time-based rules. Simply forwarding the main number is of little help if important direct extensions or service hotlines continue to go unanswered.
For critical situations, it’s advisable to provide a brief message that clearly sets expectations—for example, by promising to call back or offering an alternative way to contact the company. Here, too, less is more. When an issue arises, callers need clear information, not a lengthy explanation of the technical cause.
Check emergency calls and location information separately
Emergency call procedures must never be treated as a minor detail. With IP telephony, remote work, and mobile workstations, it must be clear which location is assigned to an extension and how emergency calls are routed. This applies in particular to 112, 117, 118, and 144 in Switzerland.
In the event of an outage, employees may need to make calls using their cell phones or work from an alternate location. Therefore, train your teams to actively provide their current location when making an emergency call. Check with your service provider and your IT department to determine what location data is technically stored and what limitations exist regarding call forwarding or softphones.
Include Microsoft Teams in the contingency plan
Microsoft Teams provides additional flexibility when employees work from home, on the go, or at another location. However, as a telephony channel, Teams also relies on identities, network access, and the underlying connectivity. Teams alone is therefore not a substitute for an emergency architecture.
Plan for Roles and Alternatives: Can a team member take calls from a queue if the reception desk is unavailable? Can call groups and availability rules be used outside the office? Is it clear how employees switch to mobile calls in the event of a WAN outage? Native integration of the PBX and Teams reduces media discontinuities and the need for separate add-on components, but it does not replace a second line or clear operational processes.
Tests determine actual availability
Even the best documentation is worthless if no one can find or understand it when an incident occurs. Therefore, test at least once a year—more frequently for critical environments. Simulate a failure of the primary Internet connection, check call forwarding, and make test calls from outside the network. Also verify that the display, voice prompts, call queue, and callback processes are functioning as intended.
A good test doesn't end with a successful call. Make a note of how long the switchover took, which employees were notified, and where manual steps were necessary. This leads to concrete improvements: an updated contact list, better labeling in the network cabinet, or a simplified call forwarding rule.
It is also important to establish clear lines of responsibility. Designate a specialist for telephony, a technical expert for the network and firewall, and a backup for each role. Store provider contact information, login credentials, and escalation procedures securely, but make them accessible to authorized personnel. Waiting on hold and unclear responsibilities cost more time during an outage than the actual technical resolution.
A solution that fits your business
Failover protection is not a one-size-fits-all solution. A single location can already be very well protected with business internet, mobile fallback, a UPS, and automatic call forwarding. Multiple locations also benefit from cross-site call groups, separate access paths, and a centrally managed security strategy. For particularly demanding requirements, redundant components, defined service levels, and regular recovery tests are added.
Winet designs these communication environments—ranging from Internet access to SIP trunking and Ayrix PBX, all the way to Teams integration and security components—as a cohesive solution. This provides a single point of contact, eliminating the need to coordinate multiple parties in the event of a malfunction.
The most sensible next step isn't a purchasing decision, but an honest failure test on paper: What would happen to your main phone number if the router, internet, or power went down tomorrow at your most critical location? If you can answer this question specifically in just a few minutes, your phone system is already much better prepared.
Current
PBX Migration: Safely Moving Your Phone System to the Cloud
How does Swiss number portability work?
Guide to Microsoft Teams Calling
Cloud Telephony for Teams That Are Always Reachable
Teams, phone service, or VoIP—which one is right for you?
