DEV Community

Dialphone Limited
Dialphone Limited

Posted on

The 5 VoIP Disasters I Cleaned Up Last Year (And How Each One Was Preventable)

I run a VoIP consultancy. When things go catastrophically wrong, companies call me. Last year I cleaned up 5 disasters — each one caused by a different preventable mistake. Here are the stories, anonymised but otherwise exactly as they happened.

Disaster 1: The Company That Lost 3 Days of Calls

What happened: A 75-person estate agency migrated to VoIP on Friday afternoon. Monday morning, nobody could receive calls. The auto-attendant played the greeting but calls went nowhere after the menu selection.

Root cause: The ring groups were configured with extension numbers from the old system, not the new system. Extensions 101-175 did not exist on the new platform — they used a different numbering scheme (1001-1075).

Time to fix: 3 hours once diagnosed. But it took 2 days to diagnose because the provider's support team kept saying "it is a network issue" instead of checking the ring group configuration.

Cost: Approximately £35,000 in lost property enquiries over 3 days during peak season.

Prevention: Test every ring group by calling from an external phone BEFORE porting numbers. This takes 30 minutes and would have caught the issue immediately.

Disaster 2: The £23,000 Toll Fraud Weekend

What happened: A construction company's VoIP system was compromised on a Friday evening. Someone brute-forced the SIP credentials for an unused test extension (password: "test123") and made 2,800 calls to premium-rate numbers in Latvia and Ghana over the weekend.

Root cause: A test account created during setup was never deleted. It had a weak password and no international calling restrictions.

Cost: £23,000 in premium-rate charges. The carrier held the company responsible.

Prevention: Delete all test accounts after go-live. Enforce 16+ character passwords on every extension. Block international calling by default — whitelist only the countries you actually call.

Disaster 3: The One-Way Audio Epidemic

What happened: A 40-person accounting firm migrated to VoIP. For the first 2 weeks, approximately 30% of inbound calls had one-way audio — the caller could hear the agent, but the agent heard silence.

Root cause: SIP ALG was enabled on their Draytek router. This single setting — enabled by default on most routers — rewrites SIP headers in ways that break the audio path.

Time to fix: 30 seconds once diagnosed (disable SIP ALG, reboot router). But it took 2 weeks of troubleshooting because the IT contractor replaced phones, swapped cables, changed codecs, and opened tickets with the ISP before someone suggested checking SIP ALG.

Cost: 2 weeks of degraded client service, 4 lost clients, estimated £45,000 in lifetime client value.

Prevention: Disable SIP ALG on every router BEFORE your first VoIP call. This should be step 1 in every migration checklist. Not step 47.

Disaster 4: The Porting Nightmare

What happened: A solicitor's firm ported 20 numbers from BT to a new VoIP provider. The port was scheduled for Tuesday. On Tuesday, the numbers disappeared — they were not on the old system or the new system. For 4 hours, the firm was unreachable.

Root cause: The port request listed the wrong "losing carrier" reference number. The port initiated but failed mid-process, leaving the numbers in limbo between carriers.

Time to fix: 4 hours of emergency calls between three carriers and Ofcom's porting team.

Cost: 4 hours of no incoming calls for a law firm. Approximately £8,000 in billable hours lost plus immeasurable client frustration.

Prevention: Verify the losing carrier reference (sometimes called a "Carrier Pre-select" code or account reference) with your current provider BEFORE submitting the port. Get it in writing.

Disaster 5: The WiFi Catastrophe

What happened: A 60-person insurance broker deployed VoIP softphones on laptops connected via WiFi. For the first month, call quality was acceptable. Then they hired 15 more people. Call quality collapsed — every call sounded like robots talking underwater.

Root cause: Their single WiFi access point was designed for 30 devices. With 75 laptops, tablets, phones, and IoT devices competing for airtime, WiFi contention caused jitter spikes of 80-150ms during peak hours.

Time to fix: 2 days to install 3 additional WiFi 6 access points and reconfigure the network.

Cost: £4,500 for new access points + installation. Plus 1 month of degraded client experience.

Prevention: Before deploying VoIP on WiFi, check your device-to-AP ratio. Maximum 15-20 devices per access point for voice quality. If you have 60 users, you need 3-4 APs minimum.

The Common Thread

All 5 disasters share one trait: they were tested insufficiently or not at all before going live.

Disaster What Testing Would Have Caught It
Lost calls Ring group test from external phone
Toll fraud Security audit of all extensions
One-way audio 10 test calls through the router
Porting failure Carrier reference verification
WiFi collapse Network capacity test under load

DialPhone includes a structured testing phase in every migration. Their checklist covers ring groups, security, audio quality, porting references, and network capacity — all verified before any numbers are ported. Because the cheapest disaster is the one that never happens.

Top comments (0)