Is Twitter's Promoted Trends And Traffic Surge Causing Fail Whale?

Twitter finally has an ad model, but some want to know if the future of advertising at Twitter could stifle the site's growth. Some advertising executives see constraints tied to the Promoted Tweets ad model, as more steps in ad options are added and technical challenges related to services not previously tried or anticipated are also slowing progress.

Recent outages have been blamed on increasing site traffic and server issues. While Twitter has been forthcoming on troubling downtime problems related to the site, this does not give comfort to some marketers relying on the platform.

To top it off, Twitter Wednesday rolled out Promoted Trends, which debuted promoting "Toy Story 3" in the Trending topic category, which represents the most talked about terms on Twitter at the moment. Promoted Trends moves Twitter into the next advertising phase, as an extension of the Promoted Tweets platform, says Carolyn Penner, company spokesperson. It shows site visitors time-, context- and event-sensitive trends promoted by advertising partners. Brands can pay to appear at the bottom of the Trending list. It's an innovative concept.

But imagine if Google told advertisers and marketers that their site's search engine optimization (SEO) campaigns must be good enough to land their company on the first page of Google.com before you can buy AdWords, explains Derek Rey, vice president of sales and marketing at Ad.ly, which supports in-stream advertising. "It becomes limiting," he told attendees at the OMMA Social conference in New York earlier this week.

Some wonder whether the site continues to grow too fast, although no one questions the brilliance of Twitter engineers to develop and roll out new services. But the company admits discovering "unexpected deeper issues" that "caused inadvertent downtime as a result of our attempts to make changes." The company says it may have to briefly pull the site down on occasion during the next two weeks to fix the problems. "From a site stability and service outage perspective, it's been Twitter's worst month since last October," the blog reads.

Site outages are typically caused by server overloads, when a Web site can't support distributed and processed content across multiple servers, or bugs being pushed into production not caught in quality assurance and test cycles, according to a programmer and Web developer who requested anonymity. "I'm surprised the engineers didn't realize server problems during the test cycle or anticipate bugs that could cause potential problems," he says. "I am pretty sure they have an internal network where they test everything before pushing out services."

Applications can operate in any of 20,000-plus different ways, depending on the sequence of events, he explains. Engineers can't possibly test all conditions. Finding issues after the app has been pushed out from production just means engineers haven't run into the bug. It doesn't mean they don't exist.

"It's the nature of the beast, especially for programs of this complexity," he says, calling the issues "emergent behavior," which occurs when several properties or modules operate in one environment. Similar to cells that make up a person, all must work in tandem to reach success.

The company says that during the next two weeks those visiting the site may encounter "relatively short planned maintenance on the site. During this time, the service will likely be taken down." Twitter admits the traffic during the World Cup games caused outage problems, but the company will make "real-time adjustments so that we can grow our capacity and avoid outages."