When a fire tore through a NorthC data center in Almere, Netherlands, on the morning of May 7, the damage extended well beyond the building.
Vonage customers lost access to the SMS infrastructure powering some of their most critical customer-facing workflows – including identity verification, outbound notifications, and two-way messaging – and many of them stayed down for over 36 hours.
Vonage has since published an explanation. In an official blog, the company confirmed that “a fire at a data center belonging to one of our partners resulted in a total loss of connectivity for the infrastructure hosted at that location, impacting some of our messaging services.”
This sounds straightforward enough, but buried in the same update is the line that should give enterprise CX leaders pause:
“The vast majority of impacted customers had service restored within several hours as we rerouted them to an alternate data center.”
If rerouting to an alternate data center was the fix, why did it take more than 36 hours to get there for a portion of customers?
And why, for a platform that enterprises depend on to deliver real-time customer communications, was a single facility capable of causing this scale of disruption at all?
Vonage Was Not Alone – But That Is Not the Point
Vonage’s SMS API, Messages API, and Verify API – the tools contact centers and CX teams use to authenticate users, trigger notifications, and manage outbound messaging – were all knocked out.
The impact was concentrated among customers sending and receiving messages outside the US, particularly across Europe and the Middle East.
A related Numbers Deactivation API issue, tied to the same incident, ran until May 11, four days after the fire broke out.
Vonage was not the only platform caught out. IBM Cloud’s AMS03 datacenter, served by the same NorthC facility, also went offline.
Reports indicate Sev 1 support tickets went unanswered for hours and, according to The Register, IBM’s status page showed no issues at all during the disruption.
Elsewhere, Utrecht University lost its network applications and websites, and bus and tram operator Transdev reported operational impact.
The scale of the fallout from a single facility was considerable.
This was not a unique Vonage failure. The NorthC fire exposed a structural vulnerability that multiple enterprise cloud providers shared simultaneously. But that does note mean that the vendor is completely blameless.
When a platform markets itself to enterprises as communications infrastructure, the expectation is that resilience has been engineered into the architecture.
A 36-hour recovery window for a subset of customers, following an incident where a failover path clearly existed, is a legitimate problem.
A Blog Post Is Not a Resilience Strategy
In response to the outage, Vonage communicated via a public-facing update that explained the cause and confirmed a recovery path.
That is more than IBM managed in the initial hours of the same incident. But publishing a blog post after the fact is not the same as having failover architecture that prevents the blog post from being necessary.
For CX and contact center leaders, the real takeaway is about dependency.
SMS sits underneath more of the customer journey than many teams consciously track, covering everything from verification flows to live service alerts.
It is the kind of infrastructure that becomes visible only when it stops working. When it does, the customer experience fails in real time and plain sight.
What Enterprises Should Be Asking Now
The questions that should now be going to every CPaaS vendor on an enterprise shortlist are not complicated:
Where is your SMS infrastructure hosted, and across how many independent facilities?
What does your failover architecture look like, and how quickly does it activate?
Vonage’s own recovery suggests the mechanism was there; the question is why it did not move faster, and for everyone.
The fire in Almere was unforeseeable. A 36-hour outage, for a platform selling itself as enterprise-grade communications infrastructure, is harder to explain away.
You can find out more about the major CPaaS players in the space, by checking out this article.
