While Europe is still making sense of an international Visa outage in early June, fintech disruptor TransferWise experienced its own outage that lasted up to 11 hours on Thursday, with website, app and card payments unavailable to the company’s 1 million-plus users.

At some point in the morning of June 21 U.K. time, the TransferWise website started showing an error message to visitors proclaiming “technical issues” — at least one Twitter user claimed to have been without service since 10 a.m. The company issued a brief Twitter statement at 12:43 confirming that the problem was affecting their app and card users as well.

At 2:07 p.m. another announcement emerged via Twitter, blaming the incident on a “power outage in our data center”, and backtracking on the extent of the issue, claiming only that the issues “may affect our website, app, and card” (our emphasis).

With the website promising a speedy fix and no notifications sent out via email or on the company’s Facebook page, users resorted to Twitter to seek information and guidance when they found their cards being rejected and access to transfers via the app and website unavailable.

The outage bears some comparison with the recent massive failure of Visa systems across Europe on June 1, which was likewise attributed to data center problems.

In that case the downtime had a rather more complex root cause — according to a detailed report (PDF) provided by Visa Europe’s CEO to the U.K. Parliament, a “very rare partial failure” in a subcomponent of the data center’s primary switch meant that the system was unable to fall back to a backup version.

Visa Europe operates twin copies of its main system in “active-active” configuration, meaning that should the master fail, the backup always has the latest data in place ready to take over. This should, in theory, mean reliable operation at all times, but it seems there is still room for glitches if something prevents this fallback from taking effect.

Visa notes in its report that it is in the process of migrating its European data operations to its central VisaNet system, which operates four pairs of active-active systems and can boast 99.9995% reliability handling 319 billion transactions in the last few years. The migration is due to complete by the end of 2018.

Of course, a relatively youthful disruptor like TransferWise doesn’t have the clout or scale requirements of a global giant like Visa, so perhaps succumbing to an occasional data center issue is inevitable, but hopefully lessons are being learned and systems hardened against future failures.

Transferwise's Twitter feed quickly filled up with complaining users, many of them stuck in awkward situations abroad where they were relying on TransferWise’s low-cost currency exchange features to fund their travels.

Eventually the service came back online, with a message at 11:19 p.m. proclaiming card functionality had been restored, followed by another two hours later confirming all systems were back to normal.

Throughout the incident, staff were busy responding to queries from concerned users, promising a speedy recovery and offering assistance with individual issues, but as of the time of writing some 24 hours after the incident started, no information has been posted to the company’s website or Facebook feed and no explanatory emails have been sent to users. TransferWise did not respond to an inquiry from PaymentsSource by deadline.

Data center issues are key to large online businesses, and server hosting facilities should be immune to power outages thanks to redundant power sources and in-house fallback power. Of course, occasional problems are inevitable, so additional redundancy on the customer side is usually advisable, with services falling back to backup systems hosted elsewhere should the main servers fail.

If this issue is indeed due to a simple power cut, it looks like TransferWise needs to expand and solidify its back-end systems to keep track with its recent surge of growth and expansion — the company recently announced integration with the French bank BCPE Group, and while a planned partnership with fellow U.K. disruptor Starling Bank foundered, there are now rumors of a new tie-in with Monzo.

It’s also important to learn from the communication problems — an active Twitter feed is crucial, but it should not be the only means of informing users of ongoing issues.

Subscribe Now

Authoritative analysis and perspective for every segment of the payments industry

14-Day Free Trial

Authoritative analysis and perspective for every segment of the industry