Every business has a version of the same devastating story. The card machine stops authorising halfway through a Saturday lunch rush. A finance team loses access to the accounting platform on the last day of the month. A support desk watches its phone system drop mid-call with a customer who was already unhappy. Nobody planned for it, nobody caused it, and for the next forty minutes nobody can do very much about it.
The problem is that an internet outage rarely announces itself. It arrives as a series of small failures that only make sense once someone checks the router, and by then the cost has already begun to accumulate. What makes internet downtime genuinely expensive is not the technical fault itself. It is that almost everything a modern business does now depends on a single connection remaining operational: payments, phones, cloud applications, stock systems, remote staff, security cameras, and even the door entry system in some buildings.
This article looks at what actually causes an internet outage, what it costs when one hits, and what practical business connectivity solutions exist to make sure the next fault becomes an inconvenience rather than a lost trading day.
What Causes an Internet Outage?
Outages are usually blamed on the provider, but the real picture is more distributed than that. Understanding where failures originate matters, because it determines which of them you can realistically protect against.
Physical infrastructure damage
The most common cause is also the least sophisticated. Fibre cuts from roadworks, construction sites, storm damage and accidental cable strikes take out connections regularly, and they are entirely outside your control. If a contractor puts a digger through a duct two streets away, your service level agreement does not make the cable reappear any faster.
Power failures
Power remains one of the largest single contributors to network incidents, accounting for a substantial share of reported outages in recent industry analysis. A local power cut takes down your router, your switches, your access points and your phones simultaneously. Even where a building has backup power, the exchange or cabinet serving it may not.
Faults further up the stack
Something interesting has shifted in recent years. As facility-level reliability has improved, failures have migrated upward into the internet’s logical layers: DNS resolution, BGP routing and CDN edges. Uptime Institute research has tracked a steady decline in data-centre outage frequency and severity for several consecutive years, while detection of internet-topology outages has climbed.
Configuration error and human factors
Analysis of decades of outage data consistently identifies human error as a contributing factor in a large majority of incidents. A firmware update applied at the wrong time, a misconfigured firewall rule, an expired certificate, and a DNS record changed by someone who no longer works there. These are not dramatic failures, but they produce identical results.
Cyber incidents
Ransomware, DDoS attacks and compromised endpoints now sit alongside software and hardware faults as leading causes of business downtime, with recent surveys placing cybersecurity issues close behind software failure as a trigger for unplanned outages.
Capacity and contention
Not every internet outage is a total blackout. A connection that has quietly been outgrown behaves like an outage during peak periods: video calls stutter, card terminals time out, and file transfers stall. The line is technically up. Operationally, it is not.
The Real Business Impact of Internet Downtime
The financial case for resilience is stronger than most businesses assume, largely because the costs are spread across categories that rarely appear on the same report.
Direct revenue loss
If customers cannot pay, they leave. For a retailer, restaurant or clinic, this is immediate and unrecoverable. Telecoms analysis in 2026 put typical direct revenue loss for small and mid-sized retail and hospitality sites in the region of £1,000 to £3,000 per hour during a full outage, with timing mattering enormously.
Payroll that keeps running
Staff continue to be paid whether or not the systems they need are available. A fifteen-person professional services firm losing an hour of productive work is absorbing several hundred pounds in wages for output that never happened, before any client impact is counted.
Recovery time, not just outage time
This is the cost most businesses underestimate. Median duration for a significant outage sits around 53 minutes according to Uptime Institute survey data, but that median conceals a long tail, with a meaningful share of incidents running beyond four hours. And the outage ending is not the same as normal service resuming. Backlogged orders, re-queued transactions, rescheduled appointments and unanswered messages all have to be worked through afterwards.
Network Failover: How Businesses Actually Stay Connected
Network failover is best achieved with the practice of running a secondary connection alongside your primary one, with equipment that detects failure and reroutes traffic automatically.
How it works in practice
A dual-WAN router continuously checks whether your primary line is still passing traffic, typically by testing for a response from a point further out on the network. When those checks fail for a defined period, the router treats the line as down and switches everything across to the backup path.
Why business mobile broadband is the practical backup
The most sensible internet backup for most sites is a mobile one. A 4G or 5G connection uses entirely separate infrastructure from your fixed line, which is exactly the point. A fibre cut in the street has no effect on a mast. Business mobile broadband delivers genuine path diversity rather than a second connection that fails for the same reason as the first, which is the common flaw in running two fixed lines from the same exchange.
What backup internet for business should be protecting
Not every system needs to run at full speed during a fault. The value of a properly configured failover setup lies in prioritising what genuinely cannot stop:
- Card payments and POS. Usually the fastest route to a measurable return on the whole arrangement.
- Hosted voice and VoIP. Modern 4G and 5G connections with correctly configured quality of service handle business calling without difficulty.
- Cloud applications. Microsoft 365, CRM, accounting and scheduling platforms are unusable offline for most teams.
- IoT, CCTV and remote sensors. Devices keep reporting through a fixed-line fault rather than going dark.
- Site-to-site VPN access. Worth checking in advance, as IP allow-listing can break during failover unless a static IP is arranged on the backup.
Outage-Proof Business Connectivity with Pentagram Systems
Everything above leads to the same uncomfortable conclusion. An internet outage is not an unusual event, it is a scheduled certainty with an unknown date, and almost none of its causes are things a business can prevent. A digger in the street, a power cut at the exchange, a routing fault thousands of miles away: the list of things outside your control is considerably longer than the list inside it.
Path diversity chosen around your sites, not around a contract
Two fixed lines sharing a duct are not redundancy; they are the same risk purchased twice. Pentagram Systems provisions business mobile broadband across all four major UK networks, and through its BT Wholesale partnership offers Complete Mobile on EE, independently ranked as the UK’s best network for over twelve consecutive years.
One partner, one point of accountability
The worst moment to discover that your broadband provider, mobile provider and IT support each consider a fault to be one of the others’ responsibility is while a customer is standing at the till with a card in their hand. Pentagram brings business broadband, mobile, IoT, voice and cloud together under a single managed platform, with one point of contact for provisioning, billing and 24/7 technical support. When a primary circuit fails, one team sees the failover event, keeps your payments and phones running on the backup path, and works the fault to resolution. There is no triage phase spent establishing whose problem it is.
Business connectivity solutions that scale across sites
For multi-site organisations, this is the difference between a resilience strategy and a collection of separate broadband contracts whose gaps only become visible under pressure. Network failover, mobile estate management, IoT connectivity and voice provisioned through one platform mean new locations inherit the same protection by default, rather than each site becoming its own small project with its own quiet vulnerabilities.
Find out what the first hour would actually cost you
Most businesses have never calculated their exposure. They know an internet outage is bad. They have not worked out what an hour of internet downtime costs in lost transactions, idle payroll and the backlog that has to be cleared afterwards, which is precisely why the internet backup arrangement that would have prevented it never quite reaches the top of the list.
It is a short conversation, and the numbers tend to be clarifying.
Talk to Pentagram Systems about a connectivity resilience review. We will map your critical systems, identify where a single fault would take your operation offline, and design a failover setup sized around what genuinely cannot stop. No obligation, and no assumption that you need more than you do.
FAQs
1. How long does a typical internet outage last?
Survey data from Uptime Institute puts the median duration of a significant outage at around 53 minutes, though a notable share of incidents run well beyond four hours. Physical damage such as a fibre cut generally takes the longest, because repair depends on locating and physically fixing the cable rather than resetting equipment. This is the main argument for failover: you cannot shorten the repair, but you can stop it from mattering.
2. Will 4G or 5G failover actually be fast enough to run my business?
For the systems that matter most during an outage, yes. Card payments, hosted voice, email and most cloud applications run comfortably over a modern mobile connection with quality of service configured correctly. What failover is not designed for is sustaining every workload at full primary-line speed indefinitely. The right approach is to size the backup around your critical systems and accept that bulk file transfers can wait.
3. Do I need to replace my existing broadband or leased line to add failover?
No. A failover router is installed in parallel with your current primary connection and monitors it. Your existing line, contract and configuration stay as they are. This is one reason internet backup is usually straightforward to add to an established site.
Leave a Reply