Innovation is our regular column that highlights emerging technological ideas and where they may lead.
Millions of conversations around the world were rudely interrupted last Thursday when a massive cyber-attack shut down the microblogging service . Normal service wasn鈥檛 resumed . Google, and blogging service also came under attack during the same period.
The incident, believed to have been the result of efforts to silence a single Georgian blogger, has demonstrated the vulnerability of social networks and the power of . But it鈥檚 also provided a real-world experiment in users鈥� reactions when a supposedly critical service is abruptly yanked away. Surprisingly, the outages provoked relatively little outrage 鈥� but why?
Advertisement
Poor performance?
The operators of most online services view downtime as something to be avoided at all costs. After all, the consequences of websites being broken or unresponsive have been well documented.
Delays of a second or more when using software (whether on- or offline) cause users to become frustrated and lose their mental connection to what they鈥檙e doing. After ten seconds, they start . If this continues, they defect to competing services, with grave consequences for the operators鈥� .
On the basis of that research, you might think Twitter (and to a lesser degree, Facebook) would be suffering. But reading the blogs, news reports and tweets over the past few days suggests otherwise. Even heavy Twitter users seem to have been remarkably unfazed by the service鈥檚 protracted disappearance.
Fun from failure
Peter Alguacil, an analyst at Swedish website monitoring company , points out that this may be because Twitter鈥檚 overall performance has never been so good, despite the outage.
Pingdom鈥檚 figures show that Twitter was up and running for 99.96% . That figure has dropped to 98.41% during 鈥� but even that would have amounted to a last year, when the site was struggling to cope with huge growth.
In addition, Twitter has somewhat inadvertently developed a novel strategy for dealing with poor performance: persuade your users that your failings are fun. The 鈥溾�� that Twitter displays during service outages has become a much-loved mascot whose appearance has become something of an in-joke. (In fact, its failure to appear last week was considered by some Twitter veterans to be the most unsettling feature of the attacks.)
Alguacil tells me that Twitter, like other successful web 2.0 operations, has experienced successive waves of poor service performance and behind-the-scenes upgrades, as explosive growth has taxed its capacity. Users still seem to be enjoying riding that rollercoaster, thanks to a degree of empathy with their fellow users, and with Twitter鈥檚 team, including its business-plan-seeking founders.
Whale of a time
As a result, Twitter and Facebook are probably better placed to avoid the consequences of downtime than more critical, but less social, sites such as Google search.
Analysts Gartner that established sites with many competitors, low switching costs and low brand loyalty are most at risk.
Google ticks the first three boxes, so if it disappears even for a short time, users will grumble and switch to an alternative like Bing, perhaps being tempted away for good. Its brand loyalty is strong, but not strong enough that people will quit searching at all for a while until it comes back.
The strong roller-coaster-riding community of Twitter, by contrast, have tied their personas to the service. They simply embraced the fail, enjoyed taking a break from maintaining their 140-character selves, and prepared to celebrate when the service came back.
The party rolls on; behind the scenes, servers are likely being upgraded. Twitter may not have a business plan yet, but last week鈥檚 experiment showed its biggest asset 鈥� its culture 鈥� is going strong.
