There are times, however, when routine maintenance is far from routine. Twitter discovered this the hard way earlier this week when scheduled maintenance to upgrade its infrastructure went terribly astray. In the process, Twitter’s goodwill took a major hit at a time when it is starting to be embraced as the hot new communications tool.
Whether Twitter fell victim to its own success or simply was unprepared is neither here nor there, although Twitter’s hosting provider, Joyent, has been released of its duties.
More important is how Twitter handled the maintenance crisis. Truth be told, it didn’t do well – leaving many users extremely, well, pissed. (Here’s a we’re-really-sorry blog post by Twitter’s Evan Williams.)
Twitter may have failed another major goodwill test but few companies really handle unexpected service outages well. Another example is Performancing Metrics, which saw a 24-hour maintenance cycle last nearly four days.
It wasn’t so much that the online statistics service was down as much as the cavalier attitude the company seemed to have for its customers. After the 24 hour-perod expired, there were no apologies, no offers to make up for it in some way, no detailed explanation about what happened. Not impressive.
There’s no lack of competition in every online market. Customers can empathize when things stop working. But if they happen too often or without a company being straightforward about what’s happening, you can lose goodwill and customers pretty quickly.
More: VentureBeat talks about how Twitter’s problems are impacting other Twitter-related companies, while Dave Winer wonders why Twitter keeps going down. TechCrunch has an intriguing post that Twitter’s problems may be linked to a move to a new host called Verio, which is owned by Japan’s NTT, which could be making an investment in Twitter.